Skip to main content
. 2017 Jun 5;27(11):1623–1632.e4. doi: 10.1016/j.cub.2017.04.046

Figure 3.

Figure 3

Melanopsin Helps Track Naturalistic Changes in Spatial Brightness

(A) An image of a typical urban scene that contained regions of similar spatial contrast but differed in mean luminance (shown by bold red rectangles). To recreate natural viewing conditions of such a scene, a small region of view was extracted, and to account for common low-amplitude eye/head movements, was projected with an on-going jitter in position (frame shifted <10° at 4 Hz).

(B–D) The projected image was rendered in different spectral mixtures over time to recreate shifting the field of view to regions of different shading: all-photoreceptors (B), melanopsin-less (C), or melanopsin-only (D). Top: changes in radiance for rods/cones (gray dashed line) or melanopsin (blue solid line), presented by transitions between stimuli occurring at intervals of 1, 5, or 10 s plotted over time. Bottom: mean ± SEM normalized firing rate of MR units (n = 10, recorded in three mice) to transitions between regions of different shading for each stimulus condition. Scale bar, 20% change in firing rate, relative to baseline.

(E and F) The correlation between firing rate and stimulus radiance was calculated for MR (n = 10) (E) and non-MR (n = 84) (F) units in each stimulus condition (Pearson’s correlation coefficient comparing the firing rate of individual neurons and the normalized radiance of the stimulus). For MR units, RM one-way ANOVA revealed a significant effect of treatment p < 0.01 (post hoc Bonferroni test compared “melanopsin-less” with “all-photoreceptor” (∗∗p = 0.002) and “melanopsin-only” (p = 0.99). For non-MR units, RM one-way ANOVA revealed a significant effect of treatment p < 0.01 (post hoc Bonferroni test compared “melanopsin-less” with “all-photoreceptor” (p = 0.16) and “melanopsin-only” (∗∗∗p < 0.001).

(G) Schematic of the change in stimulus output over time (Δ radiance).

(H and I) The correlation between firing rate and the change in stimulus was calculated for MR (n = 10) (H) and non-MR (n = 84) (I) units in each stimulus condition (Pearson’s correlation coefficient comparing the firing rate of individual neurons and the normalized Δ radiance of the stimulus). For MR units, RM one-way ANOVA revealed a significant effect of treatment p < 0.01 (post hoc Bonferroni test compared “melanopsin-less” with “all-photoreceptor” (p = 0.9) and “melanopsin-only” (∗∗∗p = 0.006). For non-MR units, RM one-way ANOVA revealed a significant effect of treatment p < 0.01 (post hoc Bonferroni test compared “melanopsin-less” with “all-photoreceptor” (p = 0.11) and “melanopsin-only” (∗∗∗p < 0.004).

See also Figure S3.