Fig. 4. Short photoperiods are defined by the expression of transcriptional repressors.

a Experimental design: SP day 28 and LP day 28 the pars tuberalis was collected at 4 hour intervals for 24 h. N = 3 per time-point. b Venn diagram showing the number of diurnal genes (significantly changing throughout the day, FDR < 0.05) in the PT in LP and SP (Supplementary Data 4). c Rose plot showing the significant diurnal genes peaking at a particular phase in LP and SP. Gray bars indicate genes shared between photoperiods and yellow indicates they are unique to a photoperiod. Around the outside of the plot the durations of light and dark the animals received is plotted in gray and black respectively, the ZT times are given. d Heatmap showing the expression profile of the genes diurnal in both SP and LP (250 genes) for each individual (n = 3 per timepoint). Data are ordered by SP peak phase. Yellow are up-regulated and blue down-regulated genes, the data are scaled. Light dark bars and ZT times are given at the bottom of the heatmap. e RNA-seq log₂ CPM plots, n = 3, individual data shown. Gray shading indicates lights off. Statistical significances are in Supplementary Data 6. f Transcription factor binding site analysis by peak phase in SP and LP. Gray and white shading show transition from dark and light. The arrows represent enrichment of either ROREs (blue), E-boxes (pink), or D-boxes (red). The direction of arrows indicate the mean expression peak of genes containing the motif in their promoter and the length of the arrow is the -log₁₀ p-value from a Rayleigh test of uniformity. g Cluster analysis of the RNA-seq data identified ZT20 SP peaking and ZT4 LP peaking gene clusters that were enriched for E-boxes. Plots shown are mean normalized log₂CPM of expression profiles to visualize cluster trends. The number of genes in each cluster is indicated in the bottom left. The broken red line represents the most representive medoid gene in the cluster.