Non-canonical Phototransduction Mediates Synchronization of the Drosophila melanogaster Circadian Clock and Retinal Light Responses

Abstract

The daily light-dark cycles represent a key signal for synchronizing circadian clocks. Both insects and mammals possess dedicated "circadian" photoreceptors but also utilize the visual system for clock resetting. In Drosophila, circadian clock resetting is achieved by the blue-light photoreceptor cryptochrome (CRY), which is expressed within subsets of the brain clock neurons. In addition, rhodopsin-expressing photoreceptor cells contribute to light synchronization. Light resets the molecular clock by CRY-dependent degradation of the clock protein Timeless (TIM), although in specific subsets of key circadian pacemaker neurons, including the small ventral lateral neurons (s-LNvs), TIM and Period (PER) oscillations can be synchronized by light independent of CRY and canonical visual Rhodopsin phototransduction. Here, we show that at least three of the seven Drosophila rhodopsins can utilize an alternative transduction mechanism involving the same α-subunit of the heterotrimeric G protein operating in canonical visual phototransduction (Gq). Surprisingly, in mutants lacking the canonical phospholipase C-β (PLC-β) encoded by the no receptor potential A (norpA) gene, we uncovered a novel transduction pathway using a different PLC-β encoded by the Plc21C gene. This novel pathway is important for behavioral clock resetting to semi-natural light-dark cycles and mediates light-dependent molecular synchronization within the s-LNv clock neurons. The same pathway appears to be responsible for norpA-independent light responses in the compound eye. We show that Rhodopsin 5 (Rh5) and Rh6, present in the R8 subset of retinal photoreceptor cells, drive both the long-term circadian and rapid light responses in the eye.

Keywords: circadian clock; cryptochrome; period; phospholipase C; phototransduction; rhodopsin; timeless.

Graphical abstract

Figure 1

norpA and cry Mutants Re-entrain More Slowly than Wild-Type (A–C) Representative actograms (left) of 16 flies for each of the genotypes and histograms (right) of the last 3 days before the shift (top) and after the shift (bottom). (A) y w. (B) norpA^P41. (C) y w;; cry^b. (D) Quantification of the position of the evening peaks for the three genotypes for each day of the experiment. While wild-type flies move their evening peak by 5 hr on the first day and 0.8 hr on the second day of the phase shift, norpA^P41 mutants need almost 2 days to move their evening peak by 5.6 hr (3 hr on the first day and 2.6 hr on the second). cry mutants shift 5.6 hr in 5 days (1.9 hr on the first day, 1.8 hr on the second, 0.7 hr on the third, 0.4 hr on the fourth, and 0.8 hr on the fifth). y w, n = 58; norpA^P41, n = 31; y w;; cry^b, n = 56. See also Figure S1.

Figure 2

Rh1, Rh5, Rh6, and Gq Contribute to norpA- and cry-Independent Behavioral Light Resetting (A–D) Representative actograms (left) of 16 flies for each of the genotypes, and histograms (right) of the last 3 days before the shift (top) and after the shift (bottom). Flies were entrained to a 2-hr ramping 12-hr:12-hr LD cycle in combination with temperature cycles of 25:16°C. After 4 days, the temperature was kept constant at 25°C, and the LD regime was shifted 6 hr. The flies were kept in these new conditions for 7 days, and afterward they were in constant darkness for 5 days. (A) norpA^P41;; cry^b. (B) norpA^P41; Rh5²; Rh6¹cry^b. (C) norpA^P41;; ss¹cry^bninaE¹⁷. (D) norpA^P41; Gq¹; ss¹cry^b. (E) Comparison of the position of the evening peak for each of the genotypes for each of the days of the experiment. While norpA^P41;; cry^b flies shift their evening peak by 5.7 ± 0.2 hr during the 7 days of the new regime, none of the other mutants are able to adapt to the new conditions (norpA^P41; Rh5²; Rh6¹cry^b, 3.0 ± 0.3 hr; norpA^P41; Gq¹; cry^b, 0.5 ± 0.4 hr; and norpA^P41;; ninaE¹⁷cry^b, 2.9 ± 0.3 hr). norpA^P41;; cry^b, n = 40; norpA^P41; Rh5²; Rh6¹cry^b, n = 55; norpA^P41; Gq¹; ss¹cry^b, n = 41; and norpA^P41;;ss¹cry^bninaE¹⁷, n = 57. Error bars represent SEM. ^∗∗∗∗p < 0.0000001. See also Figures S1A, S1B, and S3.

Figure 3

Rh1, Rh5, Rh6, and Gq Are Required for Synchronized PER Oscillations in the s-LN Clock Neurons (A and B) Representative images of the staining at ZT22 and ZT10 of the l-LNvs and the fifth s-LNv (A; fifth s-LNv marked with arrowheads) and the s-LNvs (B), with PER antibody indicated in green and PDF antibody indicated in magenta to identify the fifth PDF-negative s-LNv. Scale bars, 10 μm. (C–E) Quantification of PER expression in the l-LNv (C), fifth s-LNv (D), and PDF⁺ s-LNv (E). Note that PER oscillations in the PDF⁺ and PDF⁻ s-LNvs are synchronized to the LD cycle in y w and norpA^P41;; cry^b flies, but not in any of the other genotypes, presumably because of PER is not degraded during the day (compare PER values at ZT10 between the genotypes: PER levels in y w and norpA^P41;; cry^b s-LNvs are significantly lower compared to any of the mutants; ^∗∗∗∗p < 0.0000001 and ^∗∗∗p < 0.0001, respectively). For ZT22 and ZT10 time points, between 12 (norpA^P41;; ninaE¹⁷ss¹cry^b) and 51 (y w) brain hemispheres were imaged in 2–5 independent experiments. For ZT4 and ZT16 time points, between 6 (norpA^P41; Gq¹; ss¹cry^b) and 19 (y w) hemispheres were analyzed in 3 independent experiments (see Table S1 for exact numbers). Error bars indicate SEM. See also Figure S2 and Table S1.

Figure 4

Rh5 and Rh6 Dominate the norpA-Independent ERG Responses (A) ERG in response to a 2-s bright white light flash (bar, equivalent to ∼10⁷ effective photons per photoreceptor) in norpA^P41 (blue), Rh5²; Rh6¹ (black), and norpA^P41; Rh5²; Rh6¹ (purple) mutants. (B) ERG responses in response to a 2-s bright white flash. Top traces recorded from norpA^P41 flies (blue), norpA^P41; Rh5²; Rh6¹ flies lacking both R8 opsins (purple), and norpA^P41;; ninaE¹⁷ mutant flies lacking the R1–R6 opsin (green). Lower traces recorded from norpA^P41 flies (blue; same as upper panel), norpA^P41; Gq¹ flies (red), and norpA^P41; Gq¹/CyO flies (black; sibling controls from same vials). Each trace is a “grand” average from responses from 5 to 10 flies, each of which was itself averaged from responses to at least 4 flashes repeated at 5-min intervals. (C) Summary of response amplitudes averaged over the last 500 ms of the 2-s flash. Responses in norpA^P41; Rh5²; Rh6¹ were significantly (p = 0.005) more positive than norpA^P41 controls. Other genotypes were not significantly different (one-way ANOVA with Dunnett’s multiple comparison). (D) Summary of response amplitudes averaged between 2 and 4 s after termination of the stimulus. Responses in norpA^P41; Gq¹ recovered significantly more quickly (p = 0.002). All genotypes were also homozygous for the cry^b allele, but CRY does not influence the ERG response [36]. See also Figure S1C.

Figure 5

PLC21C Participates in Circadian Light Resetting and ERG Responses (A) Determination of Plc21C expression levels in brains and retinas of control and mutant flies by semi-qRT-PCR. Data were normalized against the ribosomal gene rp49. (B and C) Representative actograms (left) of 16 flies for each of the genotypes, and histograms (right) of the last 3 days before the shift (top) and after the shift (bottom). (B) norpA^P41; Plc21C^P319; cry^b. (C) norpA^P41; Plc21C^P319/Gq¹; cry^b/ ss¹cry^b. (D) Phase determination of evening activity peaks for norpA^P41;; cry^b (n = 40, same as in Figure 2), norpA^P41; Plc21C^P319; cry^b (n = 36), and norpA^P41; Plc21C^P319/Gq¹; cry^b/ ss¹cry^b (n = 28) flies. In the 7 days of the new light regime, neither norpA^P41; Plc21C^P319; cry^b flies nor norpA^P41; Plc21C^P319/Gq¹; cry^b/ ss¹cry^b flies are able to shift their main activity peak. (E) ERG in response to a 2-s bright white light flash in norpA^P41 (blue), norpA^P41; Plc21C^P319/+; cry^b (purple), and norpA^P41; Plc21C^P319; cry^b (green) flies. Each trace is a “grand” average from responses from 5–10 flies, each of which was itself averaged from responses to at least 4 flashes repeated at 5-min intervals. (F) Summary of response amplitudes averaged over the last 500 ms of the 2-s flash. norpA^P41; Plc21C^P319; cry^b no longer showed any hyperpolarizing component at this state (p < 0.0001, one-way ANOVA with Dunnett’s multiple comparison). norpA^P41; Plc21C^P319/+ heterozygotes also showed a reduced hyperpolarizing component, though this was only marginally significant on this sample (p = 0.06). See also Figure S1.