melanopsin expression in the brain and knockout strategy

A, B   In situs on adult brain sections reveal that (A) opn4.1 and (B) opn4xb are coexpressed in many brain domains including the epiphysis cerebri (indicated with E) or pineal. A horizontal line through a representation of the brain, above the section, indicates the location of the section within the brain. Scale bar: 1.0 mm. The brain domains in which the five melanopsins are expressed are presented in Fig EV1 and Table EV1.

C   The TALEN genome‐editing technique was applied for site‐directed mutagenesis. A 5 bp deletion was introduced proximal to the start codon of opn4.1, resulting in a frame shift and stop codon. The grey box indicates the coding sequence, the red line indicates the location of the premature stop codon and the grey triangle dispays the wild‐type sequence aligned with the mutated sequence.

D   An 8 bp deletion was introduced in the second exon (grey box) of the opn4xb gene resulting in a premature stop codon.

E   Model of the seven transmembrane G‐protein‐coupled photoreceptor shows the light absorbing molecule, the chromophore retinal (yellow), bound to helix 7 (blue). The red crosses indicate where the Melanopsins in the mutants are truncated: the mutated Opn4.1 has lost all transmembrane helices and the mutated Opn4xb ends after the second transmembrane helix. From the structure–function relationship of the photoreceptor, one can deduce that neither of the mutant Melanopsins is functional.

Expression profiling reveals pathways downstream of Opn4

A   Representation of the zebrafish adult brain shows the lateral and dorsal view with abbreviations for the following domains: olfactory bulb [OB], telencephalon [Tel], epiphysis cerebri or pineal [E], optic tectum [TeO], cerebellum [Ce], medulla oblongata [MO], hypothalamus [H] and hypophysis or pituitary [Pit]. Transcriptome sequencing was performed on cDNA from wild‐type and opn4 dko eyes and brains, sampled in the light phase (ZT4‐6). The brains were separated in an anterior part, which includes part of the forebrain and the pineal, and a posterior part, which includes part of the forebrain and the whole mid‐ and hindbrain. The cut lines through the depicted brain indicate where the brain was partitioned and which parts were subjected to transcriptome sequencing. The blue (anterior brain) and grey (posterior brain) arrows connect the brain parts in (A) with the differentially expressed genes in (B).

B  Venn diagram shows the number of differentially expressed genes in the opn4 dko anterior brain (dark blue circle), posterior brain (grey circle) and eye (light blue circle). The cutoff for differential expression was set at the significance level of α = 0.05. Heat maps of all the differentially expressed genes are presented in Appendix Fig S1A–C. The number of differentially expressed genes that are shared between the data sets is indicated where the datasets overlap.

C   Fisher's exact test shows that the shared differentially expressed genes between data sets are unlikely to be the result of coincidence. P‐values for the common differentially expressed genes are indicated where the datasets overlap. The large number of common differentially expressed genes between the data sets derived from the anterior and posterior brain parts is likely due to forebrain regions that are shared between these parts.

D   Pie chart shows the number of differentially expressed genes in the opn4 dko anterior brain that were assigned with KEGG software to phototransduction (dark blue), metabolic (orange) and other pathways. Of the metabolic pathway, 5 genes were assigned to the subcategory tryptophan metabolism (red), which encode all the enzymes that convert tryptophan into melatonin. Note that 189 genes were not assigned to a pathway.

E   Ancestor charts show the gene ontology attributes to which the differentially expressed genes in the opn4 dko anterior brain were assigned with FuncAssociate software (cutoff: α = 0.05). The number of genes assigned to an attribute is indicated in the upper right corner. All retrieved biological processes and molecular functions are associated with light, consistent with knockout of a photoreceptor.

Source data are available online for this figure.

Melatonin pathway genes are overexpressed in the opn4 dko anterior brain

A  Transcriptome sequencing reveals a significant higher number of ddc reads in the opn4 dko than in the wild‐type anterior brain, and no difference in the posterior brain and eyes. ddc reads were normalised to 1,000 actb1 reads.

B   qPCR confirms a significant higher ddc transcript level in the opn4 dko (grey box) than in the wild‐type (blue box) anterior brain. In contrast to the transcriptome data, a significant higher level of ddc was detected by qPCR in the opn4 dko eyes. Quantification of the ddc mRNA level is relative to the actb1 mRNA measured in the sample.

C   Chart shows normalised read counts for asmt. A significant higher number of asmt reads is detected in the opn4 dko than in the wild‐type anterior brain. Note that asmt mRNA is detected in the posterior brain because the pineal stalk is most likely included in this part.

D   qPCR confirms a significant higher asmt transcript level in the opn4 dko than in the wild‐type anterior brain. Quantification of the asmt mRNA level is relative to the actb1 mRNA measured in the sample.

Data information: In (A) and (C), blue markers indicate wild‐type, grey markers indicate opn4 dko and red bar indicates mean. Biological replicates are indicated with round, triangular and square markers (n = 3). Asterisks indicate significance (0.01 < P(*) < 0.05, 0.001 < P(**) < 0.01, P(***) < 0.001, ns = not significant). In (B) and (D), boxplot divides the data in quartiles: the box indicates the interquartile range, with the horizontal line in the box denoting the median of the data set, the whiskers extend to the minimum and maximum, and meet the box at the median of the lower (quartile 1) and median of the upper (quartile 3) half of the dataset. Black dots indicate biological replicates (n = 12), the red dot indicates the mean and red error bars indicate the confidence interval (95%).

opn4 dko larvae display attenuated locomotor activity in the wake state.

A-C   Actograms show mean locomotor activity of wild‐type (blue line) and opn4 dko (grey line) larvae at (A) 5 dpf, (B) 6 dpf and (C) 7 dpf. The band on each side of the mean indicates the confidence interval (95%). The bar under the chart indicates the 12:12 h LD regime.

D   Box plot shows reduced locomotor activity during the wake state in opn4 dko larvae (grey box) when compared to wild‐type (blue box) at 6 and 7 dpf. All activity bins in the light phase of each larva were pooled separately, thus the aggregated activity data of each larva on one particular day is in this case one single observation in the test statistics. As the consecutive days are presented separately, the assumption of independent identical observations is fulfiled.

E   Plot as in (D) for the rest state, shows no difference at 5 and 6 dpf and a significantly raised activity of the opn4 dko at 7 dpf.

F   opn4 dko larvae show significant more displacement in large‐angle turns at the onset of the dark interval than wild‐type on the 5^th, 6^th and 7^th dpf (see also Fig EV3A–C

G   Plot shows asmt mRNA levels, detected by qPCR, in whole wild‐type (blue box) and whole opn4 dko (grey box) larvae at 6 dpf.

H   Plot shows a significant higher asmt mRNA level in the ophthalmectomised opn4 dko larvae than ophthalmectomised wild‐type larvae at ZT3 on the 6^th dpf.

I   Plot shows similar asmt mRNA levels between opn4 dko eyes and wild‐type eyes at ZT3 on the 6^th dpf.

J   ish with asmt probe on larvae sampled on the 6^th dpf at ZT3 shows that asmt mRNA is confined to the pineal and indicates that the asmt gene is not ectopically expressed in opn4 dko larvae. The higher asmt mRNA level measured by qPCR in the opn4 dko therefore points at a defect in the pineal. Note that the eyes were removed after the ish for optimal visibility of expression in the brain. Scale bar: 0.5 mm.

K   Plot as in (G), shows significant higher levels of ddc mRNA in whole opn4 dko larvae at ZT3.

L   Plot as in (H), shows significant higher levels of ddc mRNA in ophthalmectomised opn4 dko at ZT3.

M   Plot as in (I), shows significant higher levels of ddc mRNA in opn4 dko eyes at ZT3.

N   ish with ddc probe on wild‐type and opn4 dko larvae sampled on the 6^th dpf represents ddc expression in pineals. Whole larvae are presented in Appendix Fig S2. Biological replicates indicated with BR. Scale bar: 0.05 mm.

O   Quantification of the adjusted intensity in the pineals from larvae presented in (N), implies a significant higher ddc transcript level in the opn4 dko pineal.

Data information: In (D–I), (K–M) and (O), boxplot divides the data in quartiles: the box indicates the interquartile range, with the horizontal line in the box denoting the median of the data set, the whiskers extend to the minimum and maximum, and meet the box at the median of the lower (quartile 1) and median of the upper (quartile 3) half of the dataset. Black dots indicate biological replicates (D–F: n = 18, G–I and K–M: n = 12, O: n = 8), the red dot indicates the mean, red error bars indicate the confidence interval (95%), asterisks indicate significance (0.01 < P(*) < 0.05, 0.001 < P(**) < 0.01, P(***) < 0.001, ns = not significant).

Larvae adapt their locomotor activity to the ambient light level

A–C  Actograms show mean locomotor activity in wild‐type under a medium (dark blue line) and low (light blue line) intensity LD regime at (A) 5 dpf, (B) 6 dpf and (C) 7 dpf. Spectrum and photoperiod is the same under both conditions. The band on each side of the mean indicates the 95% confidence interval. The bar under the chart indicates the 12:12 h LD interval.

D   Box plot shows a significant reduction in locomotor activity during the wake state under a low (light blue box) than a medium (dark blue box) intensity LD regime on the 5^th, 6^th and 7^th dpf.

E   As in (D) for the rest state (dark phase). No difference in locomotor activity was detected.

F   Larvae show significant less displacement in large‐angle turns at the onset of the dark interval under a low intensity LD regime than under a medium intensity LD regime on the 6^th and 7^th dpf (see also Fig EV3D–F and I).

Data information: In (D–F), boxplot divides the data in quartiles: the box indicates the interquartile range, with the horizontal line in the box denoting the median of the data set, the whiskers extend to the minimum and maximum, and meet the box at the median of the lower (quartile 1) and median of the upper (quartile 3) half of the dataset. Black dots indicate biological replicates (n = 18), the red dot indicates the mean, red error bars indicate the confidence interval (95%) and asterisks indicate significance (0.01 < P(*) < 0.05, 0.001 < P(**) < 0.01, P(***) < 0.001, ns = not significant).

The light level is correlated with per2 and asmt transcript levels

A   Per2 is known to act as an intermediary between the light detector and the melatonin synthesis pathway. Box plot shows per2 mRNA levels measured by qPCR in 6 dpf larvae placed under a low (light blue box) or medium intensity (dark blue box) LD regime. Under the low intensity light phase a significant reduction in the per2 transcript levels is detected. Bar under the chart indicates the 12:12 h LD interval.

B   Box plot as in (A) shows mRNA levels of arntl1b, which is in part regulated by per2. Under low intensity light a significant reduction in the arntl1b transcript level is detected at ZT3 and an increase at ZT21.

C   Plot as in (A) shows mRNA levels of per1a, which is regulated by the Arntl‐Clk heterodimer. Under low intensity light a significant reduction in the per1a transcript level is detected at ZT21 and an increase at ZT9.

D   Box plot shows similar per2 mRNA levels between wild‐type (blue box) and opn4 dko (grey box) under a medium intensity LD regime at 6^th dpf.

E   Box plot as in (D), shows similar arntl1b mRNA levels.

F   Box plot as in (D), shows similar per1a mRNA levels.

G   Box plot as in (D), shows similar clk1a mRNA levels.

H   Box plot as in (A) shows ddc mRNA levels. Similar levels were detected between wild‐type larvae under a low or medium intensity LD regime.

I   Box plot as in (A) shows asmt mRNA levels. Under a low intensity LD regime a significant increase in the asmt transcript level is detected at ZT3.

Data information: boxplot divides the data in quartiles: the box indicates the interquartile range, with the horizontal line in the box denoting the median of the data set, the whiskers extend to the minimum and maximum, and meet the box at the median of the lower (quartile 1) and median of the upper (quartile 3) half of the dataset. Black dots indicate biological replicates (n = 12), the red dot indicates the mean, red error bars indicate the 95% confidence interval and asterisks indicate significance (0.01 < P(*) < 0.05, 0.001 < P(**) < 0.01, P(***) < 0.001, ns = not significant).

ishs on brain sections reveal broad expression of all five zebrafish melanopsins in the mature brain

(A–E) opn4.1 (A) opn4a (B) opn4b (C) opn4xa (D) and opn4xb (E). A horizontal line through a representation of the brain, above the section, indicates the location of the section within the brain. The names and abbreviations of all brain domains in which opn4 is expressed are listed in Table EV1. Note that the duration of the colorimetric development of the ishs presented in this figure is not exactly the same. Scale bar: 1.0 mm.

ish with ddc or asmt probe on opn4 dko mature brains show wild‐type expression patterns

A, B   Same ddc expression pattern was observed in wild‐type (A) and opn4 dko (B) brains (pineal lost from the brain slices in (B) centre panels).

C, D   asmt expression in wild‐type (C) and opn4 dko (D) adult brains show that asmt is not ectopically expressed. Thus, the higher asmt transcript level measured with qPCR in the opn4 dko anterior brain can be attributed to the pineal.

Data information: The horizontal line through a representation of the brain, above the section, indicates the location of the section within the brain. Abbreviations: epiphysis cerebri (pineal) [E], caudal zone of periventricular hypothalamus [Hc], inferior raphe [IR], periventricular pretectum [Pr], preoptic area [PO], posterior tuberculum [PT], paraventricular organ [PVO], suprachiasmatic nucleus [SCN], superior raphe [SR]. Note that the duration of the colorimetric development of the ishs in this figure is not exactly the same. Scale bar: 1.0 mm.

Overview of visual motor response (VMR) data

A–C   Actograms show the mean activity of wild‐type (dark blue line) and opn4 dko (grey line) under a medium intensity LD 12:12 h regime at the light to dark boundary on the (A) 5^th dpf, (B) 6^th dpf and (C) 7^th dpf. O‐bend spike indicated with red arrowhead. The bar under the chart indicates the last 6 min of the light phase and the first 12 min of the dark phase. Solely for estimation, the mean activity over the last 6 min in the light phase is plotted as a horizontal line for wild‐type (green) and opn4 dko (red).

D–F   Actogram shows mean activity of wild‐type larvae under a medium (dark blue line) or low (light blue line) intensity LD 12:12 h regime at the light to dark boundary on the (D) 5^th dpf, (E) 6^th dpf and (F) 7^th dpf. Solely for estimation, the mean activity over the last 6 min in the light phase is plotted as a horizontal line for larvae under medium light intensity (green) and low light intensity (red).

G   opn4 dko larvae (grey box) show similar displacement in routine turns after transition from light to darkness as wild‐type (blue box) on the 6^th and 7^th dpf, but not on the 5^th dpf (U‐test, P = 0.0413). Note that displacement of each individual larva was calculated separately by comparison with its own baseline activity in the light phase. Thus the displacement of each larva on one particular day is in this case one single observation in the test statistics.

H   Wild‐type larvae show similar displacement in routine turns after transition from light to darkness under a low (light blue box) as under a medium (dark blue box) intensity LD regime on the 6^th and 7^th dpf, but not on the 5^th dpf (U‐test, P = 0.0029).

I   Plot of the photon flux at medium (dark blue line) and low (light blue line) light intensities. For biological processes, the photon flux has a higher relevance than the irradiance (power of electromagnetic radiation) because a photoreceptor is activated by a photon. The photon flux is defined as the number of photons (γ) per unit area (m²/s).

Data information: In (G) and (H), boxplot divides the data in quartiles: the box indicates the interquartile range, with the horizontal line in the box denoting the median of the data set, the whiskers extend to the minimum and maximum, and meet the box at the median of the lower (quartile 1) and median of the upper (quartile 3) half of the dataset. Black dots indicate biological replicates (n = 18), the red dot indicates the mean, red error bars indicate the 95% confidence interval and asterisks indicate significance (0.001 < P(**) < 0.01, P(***) < 0.001, ns = not significant).

Expression of genes encoding melatonin synthesis enzymes on the 6^th dpf

A   Box plot shows significant higher ddc mRNA levels measured by qPCR in whole opn4 dko larvae (grey box) at ZT3 and ZT21 (P = 0.0240) than in wild‐type (blue box). Bar under the chart indicates the 12:12 h LD interval.

B   As in (A) for the aanat2 mRNA level, shows no significant difference between wild‐type and opn4 dko whole lavae.

C   As in (A) for the asmt mRNA level, shows no significant difference between wild‐type and opn4 dko whole lavae.

D   Box plot shows aanat2 mRNA. The difference in mRNA levels between wild‐type larvae under a medium (dark blue box) or low (light blue box) intensity LD regime is not significant (ZT21: U‐test, P = 0.0684).

Data information: boxplot divides the data in quartiles: the box indicates the interquartile range, with the horizontal line in the box denoting the median of the data set, the whiskers extend to the minimum and maximum, and meet the box at the median of the lower (quartile 1) and median of the upper (quartile 3) half of the dataset. Black dots indicate biological replicates (n = 12), the red dot indicates the mean, red error bars indicate the confidence interval (95%), asterisks indicate significance (0.001 < P(**) < 0.01, P(***) < 0.001, ns = not significant).