Fig. 1

Identification of melanopsin-responsive pretectal neurons. a Spectral power distributions for cone-matched stimuli designed to strongly (Mel. High) or more weakly (Mel. Low) activate melanopsin. Lower panel shows opsin sensitivity in Opn1mw^R mice, corrected for prereceptoral filtering, and the log effective photon flux calculated for each opsin. b Example of responses of ‘sustained’, ‘transient’ and ‘OFF’ cells to Mel. High stimuli presented at ND1 (1 log unit reduction relative to a). c Scatter plot showing maximal change in firing evoked by Mel. High stimuli during first 500 ms and last 5 s of a 10-s light step for visually responsive cells identified in Opn1mw^R mice, classified according to response type (sustained, n = 72, transient, n = 121, OFF, n = 37). d, f, h Mean ± SEM normalised change in firing for Mel High and Low steps across three logarithmically spaced intensities for melanopsin-responsive sustained cell populations in Opn1mw^R (d; n = 60), Cnga3^−/− (h; n = 5) and their equivalents in Opn1mw^R; Opn4^−/− (f, n = 9). Shaded regions represent epochs of darkness. e, g, i Mean ± SEM change in firing observed during early (0–500 ms) and late components (last 5 s) of the Mel. High and Low light steps for corresponding cell populations in d, f and h. Data were analysed by two-way RM ANOVA with Sidak’s post-tests at each intensity when significant main effects of stimulus or StimulusxIrradiance were identified. *,** and *** = P < 0.05, 0.01 and 0.001 respectively, otherwise P > 0.05