Electrophysiological classification of X- and Y-cells in the cat's lateral geniculate nucleus

Abstract

Two classes of cat retinal ganglion cells--Xand Y-cells-have been described by Enroth-Cugell and Robson (1966). X-cells linearly sum the effects of stimuli presented to different parts of the receptive field, whereas Y-cells do not. For example, a “null position” can be found within the receptive field of an X-cell such that an appropriately positioned counterphased grating or bipartite field evokes no response from the cell. Y-cells do not show a null position for such stimuli (i.e. all stimulus positions within the receptive field evoke responses), and therefore these cells do not respond in a linear fashion. Shapley and Hochstein (1975) have extended this linear/nonlinear distinction to Xand Y-cells of the dorsal lateral geniculate nucleus in cats. Hoffmann, Stone and Sherman (1972) have also classified cat lateral geniculate neurons as Xor Y-cells with a battery of tests which does not include the above-mentioned test for linear vs nonlinear summation (see also Cleland, Dubin and Levick, 1971). In this classification scheme: (1) X-cells have longer response latencies to optic chiasm stimulation than do Y-cells; (2) X-cells are not excited by fast movements of a large target having a contrast opposite to that required to excite the cell through its field center (i.e. a bright target for an off-center cell and a dark target for an on-center cell), whereas Y-cells respond vigorously to such stimuli; (3) X-cells generally respond to higher spatial frequencies of squarewave gratings than do Y-cells, and X-cell responses are always modulated by the grating bars whereas Y-cells typically display unmodulated responses for higher frequency gratings; and (4) X-cells generally give a more sustained response to an appropriate standing contrast stimulus than do Y-cells. The extent to which the classification scheme of Hoffmann et al. (1972) correlates with the linear vs. nonlinear summation criteria of Enroth-Cugell and Robson (1966) has not been directly determined. In the present paper we report the results of studying cat lateral geniculate neurons, both with a battery of tests similar to that used by Hoffmann ef al. and also with a test for their spatial summation properties by a technique first described by Enroth-Cugell and Robson (1966, p. 538). We used standard, single-unit extracellular recording techniques to study 75 cells in laminae A and A, of the dorsal lateral geniculate nucleus in 9 normal cats. Varnish-insulated tungsten microelectrodes (1530 MR at 500 Hz) were used to record single-unit extracellular potentials. During the recording session, animals were anesthetized with a 70% N+300/, O2 mixture and paralyzed with a continuous infusion of 19mg/hr Flaxedil in lactated Ringers with 5% dextrose. Animals were artificially ventilated, and their end-tidal CO* maintained at 4.0”/,. The pupils were dilated with atropine sulfate, the nictitating membranes retracted with Neosynephrine, and the corneas protected with zero power contact lenses. Spectacle lenses, if needed, were chosen by retinoscopy to make the retinae conjugate with a frontal tangent screen placed 114cm from the eyes. Varnish-insulated tungsten electrodes were placed stereotaxically on either side of the optic chiasm for bipolar stimulation of retinogeniculate axons. Latency to optic chiasm stimulation for the lateral geniculate neurons was measured from the stimulus artifact to the foot of the action potential. Receptive field centers were located using small spots of light produced with a hand projector. All cells tested had their receptive fields within the binocularly viewed portion of the visual field (i.e. &45“), and 55 (73%) of the units had their receptive fields within 20” of the area centralis. In addition to measuring latency to optic chiasm stimulation, we used two additional tests utilized by Hoffmann et al. (1972). First, a cell’s response to the fast movement (200-3OO”/sec) of a large target was determined by listening to an audio representation of the cell’s action potentials. Targets were much larger than the receptive field center and extended well into the surround. For on-center cells a black disc (0.27cd/m2 on an ambient background of 6.8 cd/m’) was used, and for off-center cells a light spot (2.0 cd/m* above an ambient background of 0.51.5 cd/m*) was used. Second, a tonic vs phasic response was determined by introducing a target smaller than the receptive field center and timing the duration of the elicited response. Light spots were used for on-center cells, and black discs were used for off-center cells. Cells responding for less than 5 set were considered phasic, and cells responding for at least 15 set were considered tonic. Spatial summation of these same cells was tested by using a 9” x 9” bipartite field. Illumination of the two halves of the stimulus was sinusoidally counterphased by passing a beam of light first through two polarizing filters placed side-by-side with the axes of