Peter Gouras

Functional properties of ganglion cells of the rhesus monkey retina.

Three general classes of cells were identified in a sample of 460 cells recorded from all areas of the retina subserving the central 40 degrees of vision in the rhesus monkey. 2. One class (colour‐opponent) had sustained colour‐opponent responses and concentrically organized receptive fields, in which usually one cone mechanism mediated the centre response and one or two different cone mechanisms mediated the antagonistic surround. A few cells of this class had non‐concentric (co‐extensive) receptive field organization. 3. A second class (broad‐band) had transient responses and concentrically organized receptive fields, in which usually two cone mechanisms mediated the centre response. In most cells, the surround had the same spectral sensitivity as the centre and the cells had non‐colour opponent responses. In other cells, the surround had a spectral sensitivity different to that of the centre and the cells had colour‐opponent responses. 4. The third class (non‐concentric) did not have concentrically organized receptive fields. One group of cells had extremely phasic on‐, off‐ or on‐off responses and no spontaneous activity, another group had characteristically regular spontaneous activity and was responsive only to moving stimuli. 5. Cells of the colour‐opponent class with concentric receptive fields had the smallest centre‐sizes, which did not vary markedly from cell to cell (mean 15 mum); cells of the non‐concentric class with phasic responses had the largest centre‐sizes, which varied from cell to cell. 6. Colour‐opponent cells comprised the highest proportion of cells near the foveola; broad‐band cells comprised the highest proportion in the more peripheral areas of the retina; non‐concentric cells were equally represented in all areas.

Trichromatic colour opponency in ganglion cells of the rhesus monkey retina.

Two hundred and eleven colour‐opponent ganglion cells were studied in the central 10 degrees of the retina of the rhesus monkey, to determine the inputs which they were receiving from different cone mechanisms. Spectral‐sensitivity measurements in the presence of neutral and coloured back‐grounds showed that 24% of these cells appeared to receive input from all three cone mechanisms. 2. In 3% of the cells, the red‐sensitive cone mechanism opposed the blue‐ and green‐sensitive ones. In 18% of the cells, the blue‐sensitive cone mechanism opposed the green‐ and red‐sensitive ones. In 3% of the cells, the green‐sensitive cone mechanism opposed the blue‐ and red‐sensitive ones. 3. In 12% of the cells receiving opponent green‐ and red‐sensitive cone inputs, responses from the beta‐band of the red‐sensitive cone mechanism could be detected and distinguished from blue‐sensitive cone input. 4. All cells receiving blue‐sensitive cone input appeared to be trichromatic. The retinal distribution of cells with trichromatic input and that of cells with beta‐band responses seemed to parallel the availability of blue‐sensitive cones in the retinal area being considered. 5. The results indicate that trichromatic interactions in the macaque visual system begin in the retina.

Concealed colour opponency in ganglion cells of the rhesus monkey retina.

Criteria for distinguishing colour‐opponent from spectrally non‐opponent cells and identifying colour‐opponent subtypes on the basis of the cone inputs they receive, have been examined in ganglion cells of the macaque retina using threshold and suprathreshold stimuli, with and without chromatic adaptation. 2. Criteria based on suprathreshold responses were found to be insufficient for distinguishing between opponent and non‐opponent cells in one‐third of the sample. Criteria based on a 560 nm neutral point were found to be insufficient for distinguishing between colour‐opponent subtypes in one‐half of the remaining cells. 3. The neutral point of colour‐opponent ganglion cells varies with the geometry and intensity of the stimulus, as well as with the amount of centre‐surround interaction and the receptive‐field location of a cell. As a result, the neutral point is often an ambiguous criterion for identifying colour‐opponent subtypes on the basis of their cone inputs. 4. About one third of the colour‐opponent ganglion cells did not show colour opponency in the presence of neutral backgrounds, and only revealed this behaviour in the presence of chromatic adaptation (concealed colour opponency). 5. The proportion of these concealed colour‐opponent cells increased towards the peripheral areas of the retina.

The rhodopsin cycle in the developing vertebrate retina: I. Relation of rhodopsin content, electroretinogram and rod structure in the rat

An integrated study was made of the developmental aspects of the rhodopsin cycle in the albino rat, including measurements of rhodopsin content, electroretinogram, and photoreceptor growth. A simplified determination of rhodopsin is described which avoids isolation of the retina, or the rod outer segments. Rhodopsin was first detected 7 days after birth; it increased rapidly until 22 days, after which the concentration remained constant, while the content per eye oontinued to inorease. The electroretinogram appeared at 12 days; the amplitudes of the a- and b-waves increased rapidly until the 21st day and slowly thereafter, paralleling the rhodopsin concentration. The spectral sensitivity curve of the electroretinogram showed a maximum, coinciding with that of the difference spectrum of rhodopsin. The length of the rods and of the outer segments in both the central and peripheral areas was determined from birth to maturity. Growth of all of these structures slowed abruptly at 14 days. Closest parallelism existed between the growth of the central outer segments and the rhodopsin concentration up to 14 days. Between 14 and 22 days both the rhodopsin concentration and the electroretinogram amplitude more than doubled, although the outer segment length and total outer segment volume increased only slightly. The development of rhodopsin, the electroretinogram and the length of the photoreceptors followed typical growth curves. Consideration of these findings in the light of recent electron-microscopic observations on rod outer segments shows that rhodopsin appears concomitantly with the primitive cilium and increases rapidly with the development of the rod sacs; the electroretinogram appears only after a considerable degree of rod sac orientation has occurred.

Rod and cone signals inS-potentials of the isolated perfused cat eye

Abstract Two types ofS-potentials have been identified in the dark-adapted retina of isolated, perfused cat eyes. One type shows the action of rods exclusively, the other shows both rod and cone activity. In all of the latter units the cones make a much larger contribution to the total response than the rods and in some of these units little or no rod response can be detected at all. Although an unusual effect occurs with strong short wavelength stimulation that suggests some antagonistic receptor interaction, the results in the main support hypotheses proposing considerable independence between rod and cone signals at or before the inner nuclear layer of the retina.

Trichromatic Mechanisms in Single Cortical Neurons

By chromatic adaptation, all three cone mechanisms of rhesus monkey vision can be identified in single neurons of striate cortex. This trichromatic inter-action occurs in cells sensitive to color and indicates that striate cortical cells tend to be more wavelength discriminating than cells at lower stages of the primate visual system.

The function of the midget cell system in primate color vision

Abstract Polyak described a neural circuit in primate retina which can mediate the output of a single cone. This circuit connects a cone with a ganglion cell by either of two types of midget bipolar cells, the synapotology of which has recently been examined by electronmicroscopy. I believe that I have detected the responses of ganglion cells of this midget cell system in Rhesus monkey retina. These ganglion cells have small receptive field centers; they are excited (either at “on” or at “off”) by only one cone mechanism (either red, green or blue sensitive) in the center of their receptive field; they have small axons based on conduction velocity measurements; they are concentrated near the fovea; they respond continuously to a maintained stimulus of appropriate wavelength. Antidromic stimulation of these cells produces a distinct graded potential in the central retina. This potential can be elicited by stimulation of the parvo, but not the magno-cellular layers of the lateral geniculate nucleus implying that the midget cell system terminates in the parvo-cellular layers of this nucleus. Cells in the parvo-cellular layers tend to resemble midget ganglion cells in having only one of the three cone mechanisms mediating their receptive field center response. Striate cortical cells receive signals from several cone mechanisms in the center of their receptive field making some (those excited at “on” by one or more cone mechanisms and at “off” by another one or more) especially sensitive to color boundaries. Cells sensitive to a particular color boundary moving in one direction seem to be located next to those sensitive to the same boundary moving in the opposite direction suggesting an organization based on color contrast rather than color,per se. A model for color vision will be proposed in which six channels (based on the receptive field center mechanism of cells in the midget cell system) excite in various combinations single color sensitive cortical cells. Excitation of such a cell either at “on” or at “off” always produces the same color sensation.

Spatial summation, response pattern and conduction velocity of ganglion cells of the rhesus monkey retina

Enroth-Cugell and Robson (1966) described two classes of ganglion cells in the retina of the cat, which were independent of the on-centre/off-centre class& cation (KufHer, 1953). On the basis of responses to drifting parallel gratings with the highest spatial frequency capable of eliciting a response, the neurones were divided into Xand Y-cells: the former always showed a modulation of firing at the drift frequency. while the latter showed an unmodulated increase in the mean discharge. A number of cells of both types were also distinguished by the presence or absence of a null position for a stationary grating of low spatial frequency, at which the introduction or removal of the pattern yielded no significant response. This observation suggested that the spatial summation of X-cells was linear and that of Y-cells was non-linear (cf. Enroth-Cugell and Robson, 1966). Several reports have recently classed ganglion cells mainly on the basis of the time course of responses to maintained stimuli as well as conduction velocity (Cleland, Dubin and Levick, 1971; Fukada, 1971: Ikeda and Wright. 1972; Stone and Hoffman, 1972; Cleland, Levick and Sanderson, 1973; Cleland and Levick, 1974; Stone and Fukuda, 1974). Despite accomplished variations in nomenclature, three main classes of cells appear to have been distinguished in most of these studies. Some cells having transient responses or fast conduction velocity have been equated to Y-cells, while other cells with sustained responses or slow conduction velocity have been equated to X-cells. A third group of cells has been distinguished mostly on the basis that their conduction velocity was much slower (Stone and Fukuda, 1974) or their responses less brisk (CleLand and Levick, 1974) than those of cells equated to Xand Y-cells. The functional significance of these “W-cells” is somewhat obscured by their heterogeneity and by the controversy of whether they represent a tertium quid of the X/Y dichotomy or the basis for another dichotomy independent of the X,IY one. None of these Studies. however. have conclusively proved the postulated correspondence between classifications based on the response pattern and/or conduction velocity and that based on the linearity of the spatial summation over the receptive field. Gouras (1968. 1969) described two classes of ganglion cells in the retina of the rhesus monkey. Cells with colour-opponent properties had sustained responses (tonic) and slow conduction v-elocities, while spectrally non-opponent cells had transient responses (phasic) and fast conduction velocities. More recently. more classes and varieties of cells were described in this retina (de Monasterio and Gouras. 1975); some of these cells had colour-opponent properties but transient responses while other cells had trigger features or responses departing from the typica ones of -simple” ganglion cells. resembling the results obtained in the retina of the cat. The results reported here represent preliminary information on the linearity of the responses of macaque ganglion cells, and was directed toward a closer look of the relation between spatial summation over the receptive field, on the one hand, and reponse pattern or conduction velocity, on the other hand. Recordings were obtained in the central 20” of the retinae of adult rhesus monkeys, lightly anaesthetised with either sodium pentobarbitone (35 mg/kghr. i.v. infusion) or ketamine hydrochloride (j-20 mg/kghr. i.v. infusion), paralysed with gallamine triethiodide (15-30 mg/kyhr, i.v. infusion) and artificially respired. Rectal temperature. mean arterial pressure, ECG and expired CT& were monitored and maintained within normal values. Anaesthetic level and dosage were assessed by cortical EEG monitoring (stage II wave form) of the posterior temporal-occipital derivation; brisk reflexes but no organised responses were

Responses of macaque ganglion cells to far violet lights

Abstract In a sample of 487 colour-opponent ganglion cells recorded in the central retina of the rhesus and cynomolgus monkeys, 9% of these neurones were found to have responses with the same sign at both ends of the visible spectrum mediated by red-sensitive cones and mid-spectral responses of opposite sign mediated by green-sensitive cones. Selective chromatic adaptation showed that the responses to far violet lights (400–420 nm) were due to input from red- and not blue-sensitive cones. These responses were enhanced by backgrounds depressing the sensitivity of blue- and green-sensitive cones and they were depressed by backgrounds depressing the sensitivity of red-sensitive cones; the sensitivity of these responses was yoked to that of responses to far red lights. The relative incidence of these ganglion cells was maximal at the foveal region and decreased towards the peripheral retina. The properties of these cells are consistent with some psychophysical observations of human vision at the short wave-lengths.

Saturation of the Rods in Rhesus Monkey

ERG responses of the Rhesus monkey obtained with monochromatic flickering light subtending 90° of visual angle indicate that the time-varying rod contribution to the ERG saturates at approximately 7×109 quanta sec−1 deg−2 at 502 mμ, a value slightly greater than that required for saturation of human rod vision. The action spectrum of both the dark adapted ERG and the saturation mechanism resembles the CIE scotopic luminosity function which suggests that both processes are mediated by the rod receptors. After rod saturation, the ERG action spectrum shifts to longer wavelengths, resembling human peripheral cone vision below 550 mμ but being less sensitive at longer wavelengths. At these light levels, the action spectrum is the same whether obtained with 4 or 20 cps square-wave flicker, which indicates that the rods are no longer contributing to the response.