Spatial and temporal response properties of lagged and nonlagged cells in cat lateral geniculate nucleus

A. B. Saul, A. L. Humphrey

Research output: Contribution to journalArticle

186 Scopus citations


1. It has recently been shown that the X- and Y-cell classes in the A-layers of the cat lateral geniculate nucleus (LGN) and divisible into lagged and nonlagged types. We have characterized the visual response properties of 153 cells in the A-layers to 1) reveal response features that are relevant to the X/Y and lagged/nonlagged classification schemes, and 2) provide a systematic description of the properties of lagged and nonlagged cells as a basis for understanding mechanisms that affect these two groups. Responses to flashing spots and drifting gratings were measured as the contrast and spatial and temporal modulation were varied. 2. X- and Y-cells were readily distinguished by their spatial tuning. Y-cells had much lower preferred spatial frequencies and spatial resolution than X-cells. Within each functional class (X or Y), however, lagged and nonlagged cells were similar in their spatial response properties. Thus the lagged/nonlagged distinction is not one related to the spatial domain. 3. In the temporal domain X- and Y-cells showed little difference in temporal tuning, whereas lagged and nonlagged cells showed distinctive response properties. The temporal tuning functions of lagged cells were slightly shifted toward lower frequencies with optimal temporal frequencies of lagged X-cells averaging an octave lower than those of nonlagged X-cells. Temporal resolution was much lower in lagged X- and Y-cells than in their nonlagged counterparts. 4. The most dramatic differences between lagged and nonlagged cells appeared in the timing of their responses, as measured by the phase of the response relative to the sinusoidal luminance modulation of a spot centered in the receptive field. Response phase varied approximately linearly with temporal frequency. The slope of the phase versus frequency line is a measured of total integration time, which we refer to as visual latency. Lagged cells had much longer latencies than nonlagged cells. 5. The intercept of the phase versus frequency line is a measure of when in the stimulus cycle the cell responds; we refer to this as the intrinsic or absolute phase of the cell. This measure of response timing not only distinguished lagged and nonlagged cells well but also covaried with the sustained or transient nature of cells' responses to flashed stimuli. Absolute phase lagged the stimulus for lagged cells, led the stilulus for nonlagged cells, and approached a quarter-cycle phase lead or lag for cells that responded transiently. 6. We conclude that geniculate X- and Y-cells are distinguished by their spatial but not temporal response properties; these characteristics are largely inherited from the retina. On the other hand, lagged and nonlagged cells are indistinguishable spatially but differ temporally. These temporal differences reflect intrageniculate mechanisms and indicate that the LGN performs a major role in the temporal transformation of signals passing from retina to cortex. 7. The responses of virtually all cells to square-wave flashing stimuli could be predicted from their responses to sinusoidal stimuli by the use of response phase data and assuming linear summation. The inhibitory dip and anomalous offset discharge that characterize the responses of lagged cells to flashed stimuli are present in these linear predictions. Thus it is not necessary to invoke strong nonlinearities in the temporal domain to account for the response profiles of lagged and nonlagged geniculate cells or for sustained and transient firing patterns. 8. Lagged and nonlagged cells responded approximately one-quarter cycle apart at low temporal frequencies. Because of the latency difference between these cell types, this quarter-cycle phase difference was maintained only over a limited range of frequencies. This range roughly matched the tuning width of lagged cells, so that lagged cells tended to cease responding around the point where their responses would be a half cycle out of phase with nonlagged cells of the same center sign. We suggest that the response timing differences between lagged and nonlagged cells may be important for generating direction selectivity in visual cortex.

Original languageEnglish (US)
Pages (from-to)206-224
Number of pages19
JournalJournal of Neurophysiology
Issue number1
StatePublished - Jan 1 1990
Externally publishedYes


ASJC Scopus subject areas

  • Neuroscience(all)
  • Physiology

Cite this