Figure 2.
Melanopsin Sustains Responses to Static Images
(A) Change in firing rate of a representative MR unit as a function of location on the azimuth (at 4.5° resolution) of 13° “all-photoreceptor” (upper panel) and “melanopsin-less” (lower panel) bars (10 s every 60 s) presented for 10 s. Main plot shows change in firing over time as a heatmap (scale to right) with mean normalized change in firing at the time of peak response as a function of bar location fitted with Gaussian curve (dashed line) to left.
(B) Mean ± SEM change in firing rate for “all-photoreceptor” (black) and “melanopsin-less” (orange) bars (10 s starting at time 0) presented at RF center of MR units (n = 40). Solid lines show fit for data from 0–10 s with exponential decay curves. Separate curves were required for the two conditions (F test comparison; p < 0.0001). Transition between spectra indicated with gray shading.
(C) Mean ± SEM changes in firing as a function of bar location for a representative MR unit in first 2 s (top) and last 2 s (bottom) of 10 s of “all-photoreceptor” (black) or “melanopsin-less” (orange) stimuli. Solid lines show Gaussian fits, R2 values shown top right.
(D) Mean ± SEM firing rate (baseline subtracted) of MR (top) and non-MR units (lower panel; 18 and 51 units, respectively, recorded in 4 Opn1mwR mice) at different contrast conditions for “all-photoreceptor” (black) or “melanopsin-less” (orange) stimuli. Michelson contrast for melanopsin is shown to right in black and the mean for rod and cone opsins in gray (%). Scale bar, 5 spikes/s. Transition between spectra indicated with gray shading.
(E) A random binary modulation stimulus covering 0.1–15 Hz was generated and rendered in either “all-photoreceptor” or “melanopsin-less” stimuli. Top: example 6-s epoch of binary modulation stimulus. Bottom: power spectral density of the stimulus as a function of frequency.
(F and G) Mean ± SEM cross power spectral density (CPSD) of MR (F) or non-MR units (G) in response to binary modulation stimulus rendered in “all-photoreceptor” (black) or “melanopsin-less” (orange) spectra. Significant differences in CPSD were found <0.88 Hz in MR (2-way ANOVA comparing single unit responses; significant effect of stimulus condition; p = 0.029. Post hoc comparison between frequencies: p < 0.01 at < 0.88 Hz) but not non-MR units (2-way ANOVA finds no effect of stimulus condition; p = 0.95). Grey bar indicates frequencies over which significant differences were detected. Inset of each panel shows unique melanopsin input to MR or non-MR units (“all-photoreceptor” – “melanopsin-less” response).
(H) Histogram of the ratio of the integrated CPSD power from 0–1 Hz/1–2 Hz for individual MR (black, unfilled histogram) or non-MR units (gray filled histogram). Higher ratios indicate increased CPSD power at lower frequencies.
(I) Circles represent the locations of a simulated RF superimposed onto a natural image; RF diameter approximates that of an individual MR unit (∼13°) and is scaled assuming that each image occupies a 180° field of view. The impact of small-moderate head and eye movements was simulated by shifting RFs randomly in space in movements of 2°–20° for 50 iterations (light blue circles). The impact of changes in gaze were then modeled by shifting the RF to a random location >40° away in the scene (dark blue circles).
(J) The luminance in each RF shown in (I) (color code retained) as its location shifts according to the 50 simulated small-moderate eye movements and more substantial change in gaze.
(K) The difference in luminance (Michelson contrast) encountered with each sequential movement of the RFs during the iterations shown in (I).
(L) Probability distributions of the mean difference in luminance (Michelson contrast) over time; black distribution shows Michelson contrast for a population of MR unit RFs that tiled a number of natural images (n = 30) across 50 simulated iterations of small/moderate shifts in gaze (randomized movements <20°; simulated as in inset). Blue distribution shows Michelson contrast encountered for the same population of RFs following larger changes in gaze (random movement >40°; simulated as in inset for a population of MR unit RFs tiling 30 natural images).
See also Figures S1–S3.