(A) Dorsal view of the head of a 2 dpf embryo showing the position of the exorh positive cells of the pineal (arrow) relative to the eyes (e). (B) Embryos were raised in a 14:10 hour LD cycle, fixed starting at 50 hpf, and processed for WISH with a probe for exorh. Zeitgeiber (ZT) indicates position within the circadian cycle, with ZT0=lights on and ZT14=lights off. Dorsal views, anterior to top. Experiment was repeated three times with similar results (n≥10 embryos per time point for the experiment shown). (C) exorh mRNA levels were measured and analyzed as described in Experimental Procedures. Normalized data from three independent experiments are included, with each diamond representing the signal from one embryo (n>6 embryos per time point). White diamonds indicate lights were on (ZT2 and ZT8), and black diamonds indicate lights were off (ZT14 and ZT20). One way ANOVA, which compares all of the sample groups to each other, indicates that the means of samples taken at different points throughout the circadian cycle are not all the same (p=0.0004). Letters (a, b, c, d) denote groups of samples that are not significantly different from one another according to Tukey’s post-hoc comparison of means (p<0.05). The mean and one standard deviation for each group are indicated by the red lines overlaying the data points. To provide further information, the p values for pairwise comparisons between each of the sample groups are listed in Supplementary Table 1. Scale bars= 50µm (A) and 20µm (B).

(A) 1–4 cell stage embryos were injected with otx5 MO or otx5 MIS and then raised in a LD cycle. Time points were collected beginning at 50 hpf, and assayed by WISH for exorh expression (n≥10 embryos per time point). Experiment was repeated twice with similar results. Dorsal views, anterior to the top. Scale bar=20µm. (B) Quantification of the digital images shown in A (carried out as described in Experimental Procedures). All measurements taken from the otx5 MO samples were pooled into one set and the measurements from the otx5 MIS into a second data set. Analysis using the Independent Two Sample Student’s t-Test indicates that the signal in the Otx5 MO injected embryos is significantly lower than in the Otx5 MIS embryos (OD=0.53 ± 0.09 and 1.01 ± 0.21 for otx5 MO and otx5 MIS, respectively, p=0.0004). (C) otx5 or β-gal mRNA was co-injected with RFP mRNA into a single blastomere of 8–16 cell stage exorh:GFP embryos. Live or fixed embryos were examined at 24 hpf by fluorescence microscopy. The embryo co-injected with β-gal & RFP mRNA has no ectopic GFP expression and develops normally, with RFP expressing cells scattered throughout the embryo. The embryo overexpressing Otx5 has ectopic GFP expression that co-localizes with the RFP expression. It also has severe developmental abnormalities, as has been previously reported for Otx5 over-expressing embryos [20]. Experiment was repeated three times, and the combined results are shown in Table 1.

(A) The brains of embryos fixed 74.5 and 90.5 hpf were processed via WISH and then sectioned along the midline into left and right halves. The border of the pineal could be easily discerned in the sectioned brain (arrows). otx5 expression is localized within the pineal tissue. per3 transcripts are expressed strongly throughout the brain, including the pinealocytes, at ZT2.5 but only weakly at ZT18.5. Experiment was repeated two times with similar results, with n≥4 embryos per time point. Lateral views, dorsal to the top. (B) Embryos were injected at the 1–4 cell stage with either per3 MO or per3 MIS. Embryos were fixed at the indicated ZT and processed for WISH (n≥4 embryos per time point). (C) The OD of the WISH signal was quantified as described in Experimental Procedures. The daytime samples (ZT0, 4 and 8) from the per3 MO samples were placed into one group and the dusk/nighttime samples (ZT12, 16 and 20) into another, and then each was compared using the Student’s t-test to the analogous group from the per3 MIS samples. This demonstrated that exorh levels are significantly higher during early daylight hours (ZT 0, 4, 8) in per3 MO injected embryos compared to controls (OD 0.35 ± 0.03 and 0.87 ± 0.16 for per3 MIS and per3 MO, respectively, p=0.005). In contrast, there is no significant difference between the per3 MO and per3 MIS embryos in the evening/night time points (ZT12, 16, 20; OD = 0.94 ± 0.13 and 1.15 ± 0.07 for per3 MIS and per3 MO, respectively, p=0.08). Scale bars=20µm.

Embryos were injected with the indicated MO, raised in a 14:10 hr LD cycle, and then (A, B, E, F) fixed and processed for in situ hybridization or (C, D) analyzed by fluorescence microscopy. (A) Control embryos have high levels of exorh mRNA, while (B) expression is severely decreased in embryos injected with the exorh MO. Quantification of the in situ signal confirms that there is a significant decrease in the exorh sp MO embryos versus controls (OD is 9.19 ± 2.96 for control MO and 6.69 ± 2.39 for exorh sp MO, n≥10 embryos between 33–72 hpf, p<0.05). (C) In exorh:GFP transgenic fish, GFP is expressed in a large region that likely encompasses all pineal photoreceptors cells. (D) Transgene expression in Exorh-depleted fish is restricted to a smaller domain. (E) Control injected fish and (F) Exorh-deficient fish had indistinguishable patterns of otx5 expression in the pineal (p) and left-sided parapineal (pp). (G) The dimensions (length and width) of the fluorescent domain within the pineal were measured from digital images such as those shown in panels C and D. There is a significant difference in both dimensions between the two groups of samples (n≥20, p<0.0001 and 0.003 for length and width, respectively). (H) The dimensions (length and width) of the otx5 expression domain in control and exorh MO-injected fish were measured from digital images such as those shown in panels E and F. There was no significant change in the dimensions of the otx5 expression domain upon depletion of Exorh (n≥10 embryos, p=0.30 and 0.93 for length and width, respectively). (I) Expression of the endogenous exorh gene is high in fish injected with a control MO, and (J) much lower in fish injected with the exorh atg MO. (K) Quantification of the in situ signal indicates that expression of the endogenous exorh gene was severely reduced in exorh atg MO injected embryos compared to controls (n=10, p<0.001). (L) The length and width of the otx5 expression domain is not different between control and exorh atg MO-injected fish (n≥9 , p=0.70 and 0.58 for length and width, respectively). Experiments were repeated three times and representative images are shown. All images are dorsal views, anterior to the top. Embryos are 33 hpf (A, B), 72 hpf (C–F), and 64 hpf (I–J). Scale bars=20 µm.

(A) Embryos were injected at the 1–4 cell stage with either exorh MO or a standard control MO. Control and experimental embryos were processed in tandem for WISH (n 5≥ embryos per condition). While a separate exorh control was used for each of the aanat2 or opn1lw1 experiments, only one example is shown. For aanat2: Embryos were fixed at 2dpf ZT 18. Similar results were observed in 3 independent replicants using the exorh atg MO (shown) and 2 independent repeats using the exorh sp blocking MO (data not shown). Pineal size was confirmed through otx5 staining and no change was detected (similar to Fig 4 E,F data not shown). For opn1lw1: Embryos were fixed at 2dpf ZT 0. Results were confirmed in two independent experiments using the exorh atg MO. Dorsal views, anterior to the top. Scale bars=25µm. (B) exorh transcription is significantly lower in embryos depleted of Exorh (p=0.002, similar to Fig 4 I,J,K). Similarly, aanat2 expression is significantly reduced when Exorh is depleted (p=0.006). (C) Robust exorh expression is lost when Exorh protein levels are minimized (p<0.001). In contrast, high opn1lw1 levels are unchanged in the presence of the MO (p=0.08). The OD of the WISH signal was quantified as described in Experimental Procedures. The Two Sample Independent t-Test was used to determine significance. Each bar is the normalized value of 6 embryos total from 2 different experiments.

exorh regulation: During the day, Otx5 is present to drive exorh transcription in the pineal. However, the level of exorh expression is low due to the presence of Per3, which suppresses exorh transcription through an unknown mechanism. During the night, Per3 levels fall. Thus, Otx5 protein is free to induce high levels of exorh expression. Our results indicate that Exorh also positively influences the strength of transcription from the Exorh promoter, potentially through a multi-step signaling cascade that leads from the cell membrane to the nucleus. aanat2 regulation: Previous work demonstrates that aanat2 expression depends upon binding of Clock/BMAL and Otx5 to an enhancer region downstream of the coding region [1,3]. We have shown that Exorh protein also positively regulates aanat2 transcription. One possibility is that Exorh acts through Per2, which is also required for the light-dependent onset of aanat2 transcription in the pineal [63,64].