Acetylcholine (ACh) is one of the transmitters utilized by extrathalamic afferents to modulate stimulus-driven neurotransmission and experience-dependent plasticity in the visual cortex. Since these processes also depend on the activation of glutamatergic receptors, cholinergic terminals may exert their effects via direct modulation of excitatory neurotransmission. The objective of this study was to determine whether the ultrastructural relationships between cholinergic terminals, glutamate-immunoreactive neurons, and other unlabeled cells support this idea. Sections from aldehyde-fixed visual cortex (area 17) of adult cats were immunolabled for the following molecules: (1) choline acetyltransferase (ChAT), the acetylcholine-synthesizing enzyme; (2) L-glutamate; or (3) ChAT simultaneously with L-glutamate by combining electron-microscopic immunogold and immunoperoxidase techniques. None of the cortical terminals were dually labeled, suggesting that (1) the labeling procedure was free of chemical or immunological cross reactions; and (2) glutamate immunoreactivity probably reflects the transmitter, and not metabolic, pool of L-glutamate. Comparisons between cholinergic and noncholinergic axons revealed that (1) ChAT-immunoreactive axons formed fewer identifiable synaptic contacts within single ultrathin sections (P < 0.01 using chi-square test); and (2) more of the cholinergic axons occurred directly opposed to other terminals (P < 0.0015 by chi-square test), including 21% of which resided directly across asymmetric, axo-spinous junctions. Dual labeling showed that a third of the synaptic targets for cholinergic terminals contained detectable levels of glutamate immunoreactivity. Some of the axo-spinous junctions juxtaposed to cholinergic axons also exhibited glutamate immunoreactivity presynaptically. These observations provide ultrastructural evidence for direct, cholinergic modulation of glutamatergic pyramidal neurons within the mammalian neocortex. Prevalence of juxtapositions between cholinergic terminals and axo-spinous synapses supports the following ideas: (1) ACh may modulate the release of noncholinergic transmitters, including Glu; (2) Glu may modulate ACh release; and (3) these processes may be concurrent with cholinergic modulation of glutamatergic synapses at postsynaptic sites.

Dark-adapted rods exert a tonic suppressive influence upon cone-mediated sensitivity to rapid flicker, a phenomenon called suppressive rod-cone interaction (SRC1). However, rod dark adaptation has negligible influence upon cone-mediated thresholds measured with more usual psychophysical procedures. The present study separately examined the influences of rod light and dark adaptation upon cone-mediated sensitivity to transient increases or decreases in illumination using sawtooth flicker with rapid-on (ramp-off) or rapid-off (ramp-on) waveforms. In the parafoveal retina, cones alone were stimulated with flicker by spatially superimposing longand short-wavelength stimuli presented in counterphase and matched in scotopic illuminance. Several different adaptation procedures were used. For higher (>4 Hz) frequencies, sensitivity of cones to both waveforms is nearly identical under any condition of adaptation; sensitivity decreases as rods progressively dark adapt. A considerably different situation exists for slower frequencies (1–4 Hz). Sensitivity of cones to rapid-off flicker is appreciably greater under light-adapted conditions confirming recent observations by Bowen et al. (1989). But as rods progressively dark adapt, sensitivity of cones to rapid-off waveforms decreases considerably while sensitivity to rapid-on waveforms is much less affected; in the totally dark-adapted eye, sensitivity to both waveforms is identical. These results confirm and extend recent physiological observations in amphibian retina (Frumkes & Wu, 1990) suggesting that SRCI specifically involves responses to transient decreases in illumination.

A quantitative electron-microscopic (EM) analysis of the development of synaptic density (number of synapses/100 μm neuropil) has been done in primary visual cortex (striate, area 17) of the Old World monkey Macaca nemesthna. A comparative EM morphological study of developing synaptic contacts also was done in the same tissue. We find that a few immature synaptic contacts are present at fetal (F) 75 days either in the marginal zone, which becomes layer 1, or in the deepest portion of the cortical plate, the future layer 6. At F90–140 days synaptic contacts are found throughout the cortical plate, but their density remains higher in lower cortical layers. By F140 days synaptic density averaged for all layers (10.9) is three times higher than at F90 days. Just before and after birth, synaptic density rises very rapidly to peak at postnatal (P)12 weeks (63) and then declines slowly to reach adult values (37.7) between 2–6 years. This pattern was further tested by comparing synaptic density in layer 2 which contains the last cells generated in the striate cortex to that in layer 6 which contains the first cells generated in the striate cortex. Layer 6 contained the first synapses, and had a higher density up to F140 days (an “inside-to-outside” distribution). Synaptic density was equal in the two layers at F152 days and P2 days, but by P12 weeks synaptic density in layer 2 was 27% higher than that in layer 6 (an “outside-to-inside” distribution). After P12 weeks, the synaptic density declined 51% in layer 2 and 21% in layer 6 so that both layers achieved similar densities by P6 years.

A light and EM comparison of neuropil and synaptic contact morphology finds that, at each age up to birth, synapses in layer 2 are generally less mature than those in layer 6, but these differences disappear shortly after birth. Between P6–24 weeks, synaptic contacts throughout the cortex acquire a mature morphology that clearly differentiates between asymmetric and symmetric types, although asymmetric contacts continue to acquire more postsynaptic density until adulthood.

This complex developmental pattern suggests a sequence for synaptic developments which is more related to neuron birthdate than to the arrival of extrinsic pathways or developmental events occurring in specific laminae.

Visual tasks that are perceptually diverse might be expected to elicit unique evoked-potential waveforms that exhibit differing topographic maps. To investigate this possibility, multichannel visual-evoked potentials (VEPs) were recorded in response to several dot spatial localization stimuli that are physically similar yet produce different percepts (vernier offsets, stereoscopic disparity, bisection, orientation, and relative displacement) to determine if the unique percepts arising from these stimuli reflect the activation of different cortical neural populations. The resulting evoked potentials were all similar in waveform, although the stereoscopic VEPs were relatively delayed. Topographic maps of the evoked-potential activity to each stimulus revealed a late major component with two independent foci: one 7 or more centimeters above the inion lateral to the midline, and the other at least 6 cm lateral to Oz. The scalp localization of both peaks was independent of both the position of the stimulus in the visual field and the particular stimulus cue presented. An asymmetric response to pattern appearance vs. disappearance indicated strong pattern specificity for each stimulus type except unreferenced motion. The timing of the VEP responses and relative insensitivity to retinal locus of stimulation suggest the involvement of higher cortical areas. The two map foci might be interpreted as activation of inferotemporal and parietal cortices whose roles are thought to be visual object interpretation and spatial attention and localization, respectively.

In vitro mutagenesis and germline transformation were used to create a Drosophila mutant, ΔAsn20 lacking the N-linked glycosylation site near the amino terminus of the major rhodopsin (Asn20-Gly-Ser changed to Ile-Gly-Ser). Low opsin protein levels are detected in ΔAsn20 photoreceptors. Electroretinogram responses of mutant flies show that the residual rhodopsin found in this mutant is capable of initiating phototransduction. The organization of rhabdomeres, the photoreceptor organelle containing nearly all of the rhodopsin, is aberrant in the ΔAsn20 mutant and undergoes age-dependent deterioration. These results establish that an N-linked glycosylation site, and likely glycosylation itself, plays a critical role in the maturation of Drosophila rhodopsin.

The development of lateral inhibitory interactions in the infant visual system, as reflected by the visual-evoked potential (VEP), was studied using a radial, asymmetrical windmill-dartboard stimulus. This contrast-reversing stimulus generates VEP responses with a strong fundamental frequency component and an attenuated second harmonic component (relative to that obtained using a symmetrical stimulus). These two harmonic components reflect distinct phenomena, and appear to be the result of short-range (the fundamental) and long-range (attenuated second harmonic) lateral inhibitory interactions elicited by differential luminance-modulation of contiguous spatial regions. We studied the development of the short-and long-range interactions at 100% and 30% contrast in human infants using both VEP amplitude and phase measures. Attenuation of the second harmonic (long-range interactions) was adult-like by 8 weeks of age while the strength of the fundamental (short-range interactions) was adult-like by 20 weeks suggesting a differential development of long-range and short-range interactions. In contrast, corresponding phase data indicated significant immaturities at 20 weeks of age for both the short-and long-range components.

In normal observers, sensitivity of cones to rapid sinusoidal flicker decreases by about 0.7 log units as rods progressively dark adapt. However, Arden and Hogg (1985) described a night-vision disorder characterized by normal rod sensitivity but exaggerated suppressive rod-cone interaction (SRCI). We refer to this condition as the exaggerated SRCI syndrome (ESS). The present paper examines the influence of rod-adaptation upon cone-mediated responses to light onset and offset in an observer with ESS. Under all conditions of adaptation examined, sensitivity of cones to rapid-on waveforms is indistinguishable to that of a normal observer tested under identical circumstances; rod sensitivity is also normal. However, the sensitivity of cones to transient decreases in illumination is clearly subnormal under light-adapted conditions. This deficit in cone responsiveness to light offset becomes increasingly subnormal as rods dark adapt and, when completely dark adapted, the ESS observer is nearly blind to 1 Hz rapid-off sawtooth waveforms. These results strongly bolster previous results that suggest that suppressive rod-cone interaction is restricted to the response to transient decreases in illumination.

The turtle retina has been shown to have a variety of different morphological ganglion cell types as well as distinct physiological ganglion cell types. The major projection of the retina to the brain in nonmammalian vertebrates is to the optic tectum. In this study, we address the question of which retinal ganglion cell types project to the optic tectum in the turtle.

Fluorescent rhodamine-labeled microspheres were used to trace the retinal ganglion cell projection to the superficial layers of the optic tectum. The fluorescent ganglion cell somata, retrogradely marked by transport from the contralateral optic tectum, were impaled with micropipettes containing rhodamine-horseradish peroxidase solution and this dye was iontophoresed into the cells under visual control.

Most of the morphological ganglion cell types described in Golgi studies (Kolb, 1982; Kolb et al., 1988) were stained. Thus, the small cell types G1, G2, G3, G5, G6, and G7; the medium-sized types G10, G11, G12, G13, and G14; and the large-sized types G15, G16, G19, G20, and G21 project to the optic tectum in the turtle. We have added a new type, G2a, which proves to have some differences from the original G2 in branching pattern. We were unable to stain the small type G4, the medium-sized types G8 and G9, and the large cell types G17 and G18; this suggests that they might not project to the superficial layers of the dorsolateral optic tectum, at least, in the turtle.

We have analyzed aspects of photoreceptor topography in wholemounts of human fetal retinae in the age range 13–24 weeks of gestation. Fetal retinae were stained with cresyl violet and the sizes and packing densities of rods and cones analyzed in the conventional manner.

Cones and rods were present within a differentiating region, free of mitotic figures and approximately centered on the putative fovea, represented by the foveal cone mosaic. At 13 weeks of gestation the foveal cone mosaic was clearly differentiated, cone nuclei reaching a packing density of 14,200 per mm; a small number of rods were present in the immediately adjacent region. The packing densities of both rods and cones in these regions gradually increased and the area of the foveal cone mosaic gradually decreased throughout the age range sampled, although individual variations were evident. By 24 weeks of gestation, cone density was approximately 38,000 per mm in the foveal cone mosaic. The maximum rod density observed was 59,200 per mm in the region surrounding the foveal cone mosaic in a specimen of 20–21 weeks of gestation. In all specimens, maximum cone density occurred within the foveal cone mosaic and gradually declined towards the periphery of the differentiating region; a pronounced inverse relationship between cone soma diameter and packing density was also observed. The evidence strongly suggests that both rods and cones migrate centripetally, that is towards the center of the developing fovea, from early in development, possibly from the time that they first differentiate. The implications of these findings for fovealdevelopment are discussed.

Inferior parietal and inferotemporal cortex, which process different aspects of visual information through largely segregated pathways from the visual cortex, both receive thalamic afferents from the pulvinar complex. We examined the topography of pulvinar projections to these two cortical regions by placing multiple injections of different tracers (fluorescent dyes, horseradish peroxidase) in the inferotemporal and inferior parietal cortex of macaque monkeys.

The patterns of label observed after injections in inferotemporal gyrus indicate that area TEO and the ventral part of area V4 receive a major input from the ventral part of the lateral pulvinar (PuLv) while area TE has strong connections with the caudal pole of the medial pulvinar (PuM) and only minor connections with PuLv. In contrast, injections in the caudal inferior parietal cortex demonstrate that area PGc, on the lateral surface of the inferior parietal gyrus, and area POa, in the ventral bank of intraparietal sulcus, receive strong projections from PuM and the adjacent fringe of the dorsal part of the lateral pulvinar (PuLd).

Paired injections of two different tracers in the inferotemporal and inferior parietal cortex of the same hemisphere revealed a nearly complete segregation of the two populations of labeled neurons in the pulvinar, with only a small region of overlap in PuM, close to the PuM/PuLd border. These results demonstrate a clear separation of the thalamic afferents to the inferior parietal and inferotemporal cortex which parallels the separation of prestriate afferents to these two cortical territories (Morel & Bullier, 1990).

One type of interplexiform cell (IPC) in the dace retina was discriminated physiologically from other cell classes and identified morphologically with HRP staining. This type responded with slow hyperpolarizing potentials to white diffuse light, and in addition a slow hyperpolarization (after potential) was observed after the cessation of light with relatively high intensities. The latency of the ON phase of the response was always longer than that of the second-order neurons and the amacrine cells. Morphologically, this type of IPC was similar in appearance with the dopaminergic IPC. The conventional synaptic specialization between this type of IPC and horizontal cells was observed, and the IPC was presynaptic.

Simultaneous recording in the cat's retina and lateral geniculate nucleus (LGN) was used to find excitatory inputs to LGN cells. These recordings, correlated with measurements of LGN cell receptive-field properties, suggested new functional subdivisions of LGN cells. Distinctions between lagged and nonlagged cells were described before (Mastronarde, 1987a,b; Mastronarde et al., 1991), classification of nonlagged cells is examined here.

The Xs-type relay cells described before (Mastronarde, 1987a,b) each had detectable excitatory input from only one retinal X cell. Cells that received significant input from more than one retinal X cell were of three kinds: relay cells with pure X input (XM); relay cells with mixed X and Y input (X/Y); and cells that could not be antidromically activated from visual cortex (XI). In the series of relay cells, XS-XM-X/Y-Y, cells had progressively larger receptive-field centers, lower spatial resolution, and faster and more Y-like responses to various stimuli. XI cells resembled XM and X/Y cells in some respects but tended to have higher maintained firing rates, more sustained responses, and weaker surround suppression of the center response.

The distinctness of XS, XM, X/Y, XI, and Y from each other was examined with a modification of discriminant analysis that allowed cells to lack measurements for some parameters. Any given pair of categories could be distinguished reliably with only three parameters, although less so for X/Y-Y. In particular, XI cells were distinguishable from relay cells by properties other than the results of cortical stimulation, thus supporting the identity of XI cells as a separate class of X interneurons.

Two discontinuities in the behavior of retinal input suggest that XM cells are a separate class from XS and X/Y cells: (1) LGN X cells received either no detectable input from any of the retinal X cells adjacent to their main input, or an easily detectable amount from several such cells; and (2) cells received either no Y input or a certain minimum amount. No such discontinuity in input underlies the distinction between X/Y and Y cells.

LGN Y cells were also heterogeneous. Those with substantial input from more than one retinal Y cell had larger receptive fields and a greater preference for fast-moving stimuli than did Y cells dominated by a single input. Three Y cells could not be antidromically activated. They tended to differ from Y relay cells and resemble X interneurons in several ways. These shared properties, and the general reliability of cortical stimulation for nonlagged cells, indicate that the cells were Y interneurons.

The strength of excitatory input extrapolated to zero at a separation between LGN and ganglion cell receptive fields equivalent to the radius of a retinal X axonal arbor for X input to XM, XI, and X/Y cells, or to the radius of a Y arbor for Y input to X/Y and Y cells. Thus, a retinal axon appears to be selective in providing input primarily to cells with somata within its arbor, rather than to all cells with overlapping dendrites.

Coverage, the number of receptive-field centers overlapping a single point, was estimated for each kind of LGN cell described here. Each had a coverage of at least 6, comparable to that of retinal Y cells; most kinds had coverages of 15–35. These estimates support the idea that these subdivisions of LGN cells are functionally significant.

XM and X/Y cells fill in the functional gap that is present between retinal X and Y cells and make the distribution of spatial properties more continuous, while multiple-input Y cells broaden the range of spatial properties. One role of LGN circuitry might thus be to provide a substrate for the correspondingly broad and continuous range of spatial-frequency tuning in the visual cortex.

The inhibitory amino-acid neurotransmitter, gamma-aminobutyric acid (GABA), was localized in the pure cone retina of the lizard Anolis carolinensis by autoradiographic and immunocytochemical techniques. Uptake of [3H]-GABA labeled horizontal cells, amacrine cells, numerous cells in the ganglion cell layer, both plexiform layers, and the nerve fiber layer. Label in the inner plexiform layer showed distinct lamination.

The pattern of GABA immunoreactivity was similar to the pattern of [3H]-GABA uptake, although some differences, particularly in labeling of amacrine and ganglion cells, were observed. Immunocytochemistry revealed endogenous stores of GABA in a set of horizontal cells, amacrine cells, and cells in the ganglion cell layer. Both plexiform layers were labeled by the GABA antisera. Labeling in the inner plexiform layer (IPL) was highly stratified and GABA-immunoreactive strata were present in both sublaminae a and b. Six subtypes of conventionally placed GABA-immunoreactive amacrine cells and one displaced amacrine cell subtype were identified. Three of the six conventional amacrine cell subtypes were of pyriform morphology and three subtypes were of multipolar morphology. GABA-immunoreactive interstitial cells also were observed.

Under certain conditions the GABA antiserum labeled the cones. Etching the resin eliminated cone labeling, suggesting that GABA in the cones is present in a labile pool, unlike GABA in horizontal or amacrine cells, or the observed labeling was not due to endogenous GABA. Cones did not demonstrate [3H]-GABA uptake.

Vertebrate photoreceptors are unusual neurons in that they are capable of continuous calcium-mediated release of neurotransmitter (Trifonov, 1968; Hagins et al., 1970). In this study, we have examined the development and characteristics of calcium currents in chick cone cells placed in culture on embryonic day 8. Cone cells were identified by their lectin-binding properties, rhodopsin-like immunoreactivity, and the presence of an oil droplet. Using the whole-cell patch-clamp method, we have seen calcium currents in these cells after three days in culture, slightly before the appearance of synapses (Gleason & Wilson, 1989). Because cone calcium currents are blocked by cadmium and nifedipine but are enhanced by Bay K 8644, they most closely resemble L-type current (Nowycky et al., 1985). An unexpected feature of these currents is that their gating ranges varied widely between cells so that some cells showed the foot of their activation range at —70 mV and others as positive as —25 mV. Calcium imaging of fura-2 loaded cells was used to confirm the time course of calcium current development and describe the distribution of cytosolic calcium. As expected, depolarization of young cells failed to increase cytosolic calcium but in older cells an increase of threefold to fourfold was usually observed. Both at rest and during depolarization, most cone cells showed regional differences in internal calcium concentration. In the most mature cones, depolarization strongly elevated cytosolic calcium at the terminal end of the cell while producing a lesser change around the oil droplet and the ellipsoid region, suggesting that calcium channels are localized to the terminal.

We have previously demonstrated that the neonatal rabbit retina contains a larger complement of cells that accumulate [3H]-GABA than does the adult. In order for these neurons to be classified as GABAergic, they must also contain endogenous GABA. We now report that these same neonatal cell populations are also immunoreactive to GABA antisera. In frozen sections from rabbit retina, treated with GABA antisera, immunoreactive processes in both synaptic layers were observed at postnatal day 1. The appearance of immunofluorescent fibers precedes that of photoreceptor and bipolar cell terminals in the outer plexiform layer and is diminished by postnatal day 5. Also noted, was a 50% decrease in the density of GABA-immunoreactive cell bodies in the inner nuclear and ganglion cell layers, accompanied by an increase in cell volume and a shift toward a more spherical cell shape of the remaining cells. At postnatal day 1 and 3, we also observed immunoreactive cells having the characteristic morphology of interplexiform cells. This cell type sends branches to both the outer and inner plexiform layers, thus a morphological basis for communication between the two developing plexiform layers is present as early as postnatal day 1. Thus, retinas from neonatal rabbits have a larger complement of cells that stain for endogenous GABA than does the adult. These results coupled with our previous studies suggest that GABAergic properties are expressed by a larger number of cell types in the neonate than in the adult and are consistent with the general hypothesis that GABA functions as a trophic agent during development.

The cholinergic amacrine cells of the rabbit retina may be labeled with [3H]-Ch and the activity of the cholinergic population monitored by following the release of [3H]-ACh. We have tested the effect of muscimol, a potent GABAA agonist, on (1) the light-evoked release of ACh, presumably mediated via bipolar cells, which are known to have a direct input to the cholinergic amacrine cells and (2) ACh release produced by exogenous glutamate analogs that probably have a direct effect on cholinergic amacrine cells. Muscimol blocked the light-evoked release of ACh with an IC50 of 1.0 μM. In contrast, ACh release produced by nonsaturating doses of kainate or NMDA was not reduced even by 100 μM muscimol. Thus, we have been unable to demonstrate a direct effect of GABA on the cholinergic amacrine cells.

GABA antagonists, such as picrotoxin, caused a large increase in the base release and potentiated the light-evoked release of ACh. Both these effects were abolished by DNQX, a kainate antagonist that blocks the input to cholinergic amacine cells from bipolar cells. DNQX blocked the effects of picrotoxin even when controls showed that the mechanism of ACh release was still functional. Together, these results imply that the dominant site for the GABA-mediated inhibition of ACh release is on the bipolar cell input to the cholinergic amacrine cells. This is consistent with previous anatomical and physiological evidence that bipolar cells receive negative feedback from GABA amacrine cells.

Using suction electrodes, photocurrent responses to 100-ms saturating flashes were recorded from isolated retinal rods of the larval-stage tiger salamander (Ambystoma tigrinum). The delay period (Te) that preceded recovery of the dark current by a criterion amount (3 pA) was analyzed in relation to the flash intensity (If), and to the corresponding fractional bleach (R*0/Rtot) of the visual pigment; R*0/Rtot was compared with R*s/Rtot the fractional bleach at which the peak level of activated transducin approaches saturation. Over an approximately 8 In unit range of If that included the predicted value of R*s/Rtot, Te increased linearly with In If. Within the linear range, the slope of the function yielded an apparent exponential time constant (TC) of 1.7 ± 0.2 s (mean ± S.D.). Background light reduced the value of Tc measured at a given flash intensity but preserved a range over which Tc increased linearly with In If; the linear-range slope was similar to that measured in the absence of background light. The intensity dependence of Tc resembles that of a delay (Td) seen in light-scattering experiments on bovine retinas, which describes the period of essentially complete activation of transducin following a bright flash; the slope of the function relating Td and In flash intensity is thought to reflect the lifetime of photoactivated visual pigment (R*) (Pepperberg et al., 1988; Kahlert et al., 1990). The present data suggest that the electrophysiological delay has a similar basis in the deactivation kinetics of R*, and that Tc represents TR* the lifetime of R* in the phototransduction process. The results furthermore suggest a preservation of the “dark-adapted” value of TR* within the investigated range of background intensity.

We examined inward rectification in Limulus ventral photoreceptors using the two-microelectrode voltage clamp. Hyperpolarization in the dark induced an inward current whose magnitude was distinctly dependent on extracellular K+ concentration, [K+0]. The [K+0] dependence resembled the characteristic [K+0] dependence of other inward rectifiers. The inward current was not dependent on extracellular Ca2+ or Na+, and it was unaffected by intracellular injection of Cl−. The hyperpolarization induced currents had two phases, an early nearly instantaneous phase and a slowly developing late phase. The currents were sensitive to extracellular barium and cesium. In voltage-pulse experiments, the magnitudes of the inwardly rectifying currents were variable from cell to cell, with some cells exhibiting negligible inward currents. Large hyperpolarizations (to membrane potentials more negative than about – 140 mV) caused unstable inward current recordings, irreversible desensitization, and irreversible elevation of intracellular Ca2+ concentration. The inward rectifier provides negative feedback by tending to depolarize the cell (with inward current) in response to hyperpolarization. We suggest that the inward rectifier reduces the amount of hyperpolarization that would otherwise be generated by electrogenic processes. This feature would restrict the dynamic voltage range of the photoreceptors at very hyperpolarized potentials.

We have investigated the effects of a serotonin 5-HT2 antagonist and a 5-HTIA agonist on horizontal and ganglion cells in the rabbit retina. Simultaneous intracellular horizontal cell and extracellular ganglion cell recordings were obtained from a superfused in vitro rabbit eyecup preparation and the effects of bath applied drugs on these cells' light responses observed. Sinusoidally modulated current was also injected into horizontal cells while the extracellular spike activity of nearby, single-unit ganglion cells was monitored. Although the ON components of the light-evoked responses of ganglion cells were reduced by the 5-HT2 antagonist or the 5-HTIA agonist, the membrane potential and the light responses of horizontal cells and the 6-wave of the ERG were simultaneously unaffected. However, the drugs blocked current-driven ganglion cell spike activity induced by current injections into nearby horizontal cells. These results are discussed with respect to the site of action of these serotonin drugs and with respect to the circuitry of serotonergic neurons.