A) Schematic representation of the predicted 3D structure of RH7 (GH14208p) based on AlphaFold (Jumper et al., 2021). The structure is color-coded using a continuous spectral gradient ranging from blue (N-terminus) to red (C-terminus). The three intracellular loops are highlighted and labeled in numbered circles: (1) ICL1 with the LRTPXN motif, (2) ICL2 with the highly conserved DRY motif, and (3) ICL3, which lacks the typical QAKKMNV motif. The original illustration of AlpfaFold is shown in Supplementary Material S2A. B) Left: Schematic representation of the Drosophila melanogaster optical system. The presumed Rh7-expressing cells, according to Davis et al., (2020) are shown in addition to retinal receptor cells (black). Right: Heatmap of relative gene expression across Drosophila optic lobe cell types. TPM values were obtained by TAPIN sequencing (Davis et al., 2020) and normalized per gene (100 % = maximum TPM per gene). Color scale: white (0 %), light gray (∼50 %), black (100 %). Genes are shown on the x-axis, and cell types on the y-axis.

Fly lines

Primers

Antibodies used for immunohistochemistry

A) Phylogenetic tree of 280 r-opsin proteins, highlighting the clades containing Drosophila rhodopsins: RH1, RH2, and RH6 (blue); RH3, RH4, and RH5 (purple); and RH7 (red/dark red). Arthropod RH7 homologs are shown in red, while homologs from tardigrades and onychophorans are shown in dark red. Drosophila rhodopsins are marked with stars in their respective clade colors. B) Proposed evolutionary timeline of Drosophila rhodopsins. RH7 likely emerged in the last common ancestor of panarthropods (red circle), while the other two rhodopsin clades evolved later within arthropods (purple and blue circles). C) GsX assay showing that RH7 robustly activates the Gαz chimera in a light-dependent manner, with minimal differences between 9-cis-retinal and all-trans-retinal. No activation was detected with the Gαq chimera.

Rh70 mutants show reduced activity in the dark.

Average activity profiles of wild-type (WT) control flies and Rh70 mutants are shown for A) "normal days" (12 h light (30 µW/cm2): 12 h dark, 25 °C), B) "hot days" (12 h light (30 µW/cm2): 12 h dark, 24-29 °C), and C) "short days" (8 h light (30 µW/cm2) : 16 h dark, 25 °C). Wild-type control flies exhibit anticipatory activity in the morning and a short startle response in the evening, both are reduced in Rh70 mutants. On "short" and "hot" days, control flies shift more activity to the dark phase, while Rh70 mutants remain mainly active in the light phase. A*, B*, and C* show total activity and activity in light and dark phases. Total activity is similar on "normal" and "hot days", but Rh70 mutants are less active on "short days". Rh70 mutants are significantly more active in the light phase and less active in the dark phase than wild-type controls (****p <0.0001, *p <0.05, Mann-Whitney U-Test, R-studio).

A) The average daily activity of fly strains was recorded under a 12 h light (30 µW/cm2) : 12 h dark cycle at 25 °C. During the light phase, the light was turned off for 0.5 h every 1.5 h to observe the startle response to light-off. To compare fly lines with different baseline activity, activity was normalized to their light-phase activity. All fly strains show a startle response to each dark pulse, but this is reduced in Rh70 and per01;;Rh70 mutants. The light-off startle response varies by time of day but stays constant in per0 and per0;;Rh70 mutants. B) The startle response 4-hour assay. WT controls, as well as the heterozygous flies crossed with Rh70 or Rh7MiMIC, show a strong startle response which is characterized by a short-term increase in activity. In the homozygous Rh70 and Rh7MiMIC lines, as well as in the line carrying on one chromosome the Rh70 mutation and on the other the Rh7MiMIC mutation, this startle response is significantly reduced (***p <0.001, Mann-Whitney U-Test, R-studio).

A) Relative qPCR analysis revealed strong Rh7 expression in the lamina, followed by the brain, with minimal background expression in the retina. B) Rh7 in situ HCR in wild-type (WT) controls revealed strong Rh7 signals in the lamina (LA) and medulla (ME), and weaker signals in the central brain (CB). C) Rh70 mutants showed no detectable Rh7 in situ HCR signal, confirming the specificity of the probe and the loss of Rh7 expression in the mutant. D) Rh7-Gal4-driven UAS-GFP expression revealed robust labeling in the optic system, particularly in the lamina and medulla, as well as in various brain cells. E) Rh7-Gal4-driven MCFO labeling identified lamina neurons L5 and the medulla neuron Mi1. Right: schematic illustration of the detected neurons in the optic lobe. F & G) in situ HCR in Rh7-Gal4>GFP flies revealed that all GFP-labeled cells in the lamina (F) and medulla (G) also showed Rh7 transcript signal, confirming Rh7 expression in these neurons. H) Immunostaining of Rh7-Gal4>UAS-GFP flies with anti-GFP and anti-PER antibodies revealed co-labeling in clock neurons, suggesting that a subset of circadian neurons expresses Rh7. Open arrows mark DN1p and arrows mark DN2; double arrows mark DN1a. Scale bar in B, C, D represents 100 µm, in E, F, G 25 µm, in F’, H 10 µm and in F’’ 1 µm.

A & B) Normalized average daily activity of Darkfly and rescue fly lines recorded under a 12 h light (30 µW/cm2) / 12 h dark cycle at 25 °C. During the light phase, lights were turned off for 30 min every 1.5 h. A) Darkfly flies exhibit an enhanced startle response to light-off. B) No-rescue controls (Rh7-Gal4>UAS-empty; Rh7⁰) display a phenotype similar to Rh7 mutants, whereas the Rh7 rescue line (Rh7-Gal4>UAS-Rh7; Rh7⁰) resembles WT controls. The Rh7Darkfly (Rh7-Gal4>UAS-Rh7Darkfly; Rh7⁰) rescue shows an enhanced startle response similar to Darkfly, particularly during the second half of the day. C) To measure the startle response, the 4-hour assay was performed. The Rh7 rescue line restored the reduced startle response seen in the no-rescue controls, which behave like Rh7⁰ mutants (p < 0.0001). The Rh7Darkfly rescue showed an even stronger startle response, significantly higher than the Rh7 rescue line (p < 0.001). D) Optomotor response assays show that no-rescue controls behave like Rh7⁰ mutants, with no significant difference between them. Rh7 rescue flies respond like WT controls. Both Rh7⁰ mutants and no-rescue controls differ significantly from WT and Rh7 rescue flies (p < 0.01). Darkfly and Rh7Darkfly rescue flies do not differ from each other but show significant differences compared to WT and Rh7 rescue flies (***p < 0.001, **p < 0.01, *p < 0.05, Mann-Whitney U-Test).

A) Protein lengths of Drosophila and human opsins are shown alongside randomly selected non-opsin GPCRs. General GPCRs are significantly longer, primarily due to extended N- and C-termini, whereas opsins are shorter and largely confined to the transmembrane core. RH7 and OPN4 (melanopsin) display intermediate protein lengths relative to both groups. The red bars represent the N- and C-termini of the proteins, and the black bars represent the core protein, spanning from the 1st to the 7th transmembrane domain. Additional information, including full protein names and UniProt IDs, is provided in Supplementary Material S2G. B) Left: Predicted 3D structure of RH7 based on AlphaFold, representing a closed conformation. Center: A putative open conformation that may result from light-induced conformational changes and subsequent C-terminal phosphorylation (magenta), potentially promoting G-protein binding, as suggested by Maggio et al. (Maggio et al., 2023). Right: Hypothetical structure of RH7 in the Darkfly variant, possibly stabilized in an open, constitutively active state, independent of light.