A Cryptochrome adopts distinct moon- and sunlight states and functions as sun- versus moonlight interpreter in monthly oscillator entrainment
l-cry mutants show higher spawning synchrony than wild-type animals under non-natural light conditionsIn order to test for a functional involvement of L-Cry in monthly oscillator function, we generated two l-cry mutant alleles (Δ34 and Δ11bp) (Fig. 1a) using TALENs28. In parallel, we generated a monoclonal antibody against Platynereis L-Cry. By testing mutant versus wildtype worms with the anti-L-Cry antibody in Western blots (Fig. 1b) and immunohistochemistry (Fig. 1e–j), we verified the absence of L-Cry protein in mutants. Furthermore, we confirmed that the staining of the antibody in wildtype worms (Fig. 1e–h) matches the regions where l-cry mRNA is expressed (Fig. 1d). These tests confirmed that the engineered l-cry mutations result in loss-of-function alleles. In turn, they validate the specificity of the raised anti-L-Cry antibody.Fig. 1: l-cry–/– mutants are loss-of-function alleles.a Overview of the l-cry genomic locus for wt and mutants. Both mutant alleles result in an early frameshift and premature stop codons. The Δ34 allele has an additional 9 bp deletion in exon 3. b Western Blots of P. dumerilii heads probed with anti-L-Cry antibody. In the context of further investigations such Western blots of mutant versus wild types have been performed more than 10 times with highly consistent results. Also see further analyses in this manuscript and ref. 36. c overview of P. dumerilii. d whole mount in situ hybridization against l-cry mRNA on worm head. ae, anterior eye; pe, posterior eye. e–j Immunohistochemistry of premature wild-type (e–h) and mutant (i, j) worm heads sampled at zt19/20 using anti-L-Cry antibody (green) and Hoechst staining (magenta), dorsal views, anterior up. e, f: z-stack images (maximal projections of 50 layers, 1.28 µm each) in the area highlighted by the rectangle in (d), whereas (g–j) are single layer images of the area highlighted by the white rectangles in (e, f). In the context of further investigations such stainings of mutant versus wild types have been performed more than 10 times with highly consistent results. Also see further analyses in this manuscript and ref. 36.Full size imageWe next assessed the circalunar maturation timing of wild types and l-cry mutant populations in conventional culture conditions, i.e. worms grown under typical indoor room lighting (named here artificial sun- and moonlight, Supplementary Fig. 1b).We expected either no phenotype (if L-Cry was not involved in circalunar clock entrainment) or a decreased spawning precision (if L-Cry was functioning as moonlight receptor in circalunar clock entrainment). Instead we observed an increased precision of the entrained worm population:We analyzed the maturation data using two statistical approaches, linear and circular statistics. We used the classical linear plots5 and statistics to compare the monthly spawning data distribution (Fig. 2a–c, i). This revealed a clear difference between mutant animals, which exhibited a stronger spawning peak at the beginning of the NM phase, compared to their wildtype and heterozygous counterparts (Fig. 2a–c, Kolmogorov–Smirnov test on overall data distribution, Fig. 2i).Fig. 2: L-Cry shields the circalunar clock from light that is not naturalistic moonlight.a–d, j Spawning of l-cry +/+ (a), l-cry +/– (Δ34) (b) and l-cry −/−(Δ34/ Δ34) (c) animals over the lunar month in the lab with 8 nights of artificial moonlight (a–c), under natural conditions in the sea (d, replotted from ref. 34,50,) and in the lab using naturalistic sun- and moonlight (j, 8 nights moonlight). e–h, k Data as in (a–d, j) as circular plot. 360° correspond to 30 days of the lunar month. The arrow represents the mean vector, characterized by the direction angle µ and r (length of µ). r indicates phase coherence (measure of population synchrony). p-values inside the plots: result of Rayleigh Tests. Significance indicates non-random distribution of data points. The inner circle represents the Rayleigh critical value (p = 0.05). i–l Results of two-sided multisample statistics on spawning data shown in (a–h, j, k). The phase differences in days can be calculated from the angle between the two mean vectors (i.e. 12°= 1 day).Full size imageWe then analyzed the same data using circular statistics (as the monthly cycle is repeating, see details in Methods section), which allowed us to describe the data with the mean vector (defined by the direction angle µ and its length r, shown as arrows in Fig. 2e–g). The phase coherence r (ranging from 0 to 1) serves as a measure for synchrony of the population data. As expected for entrained populations, all genotypes distributed their spawning across a lunar month significantly different from random (Fig. 2e–g, p values in circles, Rayleigh’s Uniformity test29). In line with the observed higher spawning peak of the l-cry−/− mutants in the linear plots, the circular analysis revealed a significant difference in spawning distribution (Mardia–Watson-Wheeler test, for details see Methods section) and higher spawning synchrony of mutants (r = 0.614) than in wild types and heterozygotes (r = 0.295 and r = 0.222) (Fig. 2i). The specificity of this phenotype of higher spawning precision for l-cry homozygous mutants was confirmed by analyses on trans-heterozygous l-cry (Δ34/Δ11) mutants (Supplementary Fig. 2), and by the fact that such a phenotype is not detectable in any other light receptor mutant available in Platynereis (r-opsin130: Supplementary Fig. 3a, b, e, f, i; c-opsin131: Supplementary Fig. 3c, d, g, h, i, Go-opsin: refs. 32, 33).The higher spawning synchrony of l-cry mutants under artificial light mimics the spawning precision of wild-type at its natural habitatThis increased spawning precision of l-cry mutants under artificial (but conventional indoor) laboratory light conditions let us wonder about the actual population synchrony of the worms under truly natural conditions. The lunar spawning synchrony of P. dumerilii at the Bay of Naples (the origin of our lab culture) has been worked on for more than 100 y. This allowed us to re-investigate very detailed spawning data records from the worms’ natural habitat published prior to environmental/light pollution. For better accessibility and comparability we combined all months and replotted the data published in 192934 (Fig. 2d, h, I; see details in Methods section; r = 0.631). This analysis revealed that the higher spawning synchrony in l-cry–/– worms mimics the actual spawning synchrony of P. dumerillii populations in their natural habitat34 (compare Fig. 2c, g with 2d, h.)Given that recent, non-inbred isolates from the same habitat as our lab inbred strains (which is the same habitat as the data collected in ref. 34) exhibit a broad spawning distribution under standard worm culture light conditions (which includes the bright artificial moonlight)35, we hypothesized that the difference in spawning synchrony between wildtype laboratory cultures and populations in their natural habitat is caused by the rather bright nocturnal light stimulus typically used for the standard laboratory culture (Supplementary Fig. 1a vs. b).Lunar spawning precision of wild-type animals depends on naturalistic moonlight conditionsWe next tested the resulting prediction that naturalistic moonlight should increase the spawning precision of the wildtype population, using naturalistic sun- and moonlight devices we specifically designed based on light measurements at the natural habitat of P. dumerilii31 (Supplementary Fig. 1a, c). We assessed the impact of the naturalistic sun- and moonlight (Supplementary Fig. 1a, c) on wildtype animals, maintaining the temporal aspects of the lab light regime (i.e. 8 nights of “full moon”). Indeed, merely adjusting the light intensity to naturalistic conditions increased the precision and phase coherence of population-wide reproduction: After several months under naturalistic sun- and moonlight, wildtype worms spawned with a major peak highly comparable to the wildtype precision reported at its natural habitat (Fig. 2d, h vs. j, k), and also exhibited an increased population synchrony (r = 0.398 compared to r = 0.295 under standard worm room light conditions). This increased similarity to the spawning distribution at the natural habitat (“Sea”) is confirmed by statistical analyses (Fig. 2l): The phase difference (angle between the two mean vectors) is only one day (corresponding to 12°). In contrast, the spawning distribution of wild types under standard worm room light versus naturalistic light conditions is highly significantly different in linear and circular statistical tests and has a phase difference of 7.7 days (Fig. 2l).These findings show that it is the naturalistic light that is critical for a highly precise entrainment of the monthly clock of wild-type worms. Given that l-cry–/– animals reach this high precision with the artificial light (i.e. standard lab light) implies that in wildtype L-Cry blocks artificial, but not naturalistic full-moonlight from efficiently synchronizing the circalunar clock. This block is removed in l-cry–/– animals, leading to a better synchronization of the l-cry–/– population. This finding suggests that L-Cry’s major role could be that of a gatekeeper controlling which ambient light is interpreted as full-moonlight stimulus for circalunar clock entrainment.
l-cry functions as a light signal gatekeeper for circalunar clock entrainmentA prediction of this hypothesis is that mutants should entrain better to an artificial full-moonlight stimulus provided out-of-phase than their wild type counterparts (in which L-Cry should block the “wrong” moonlight at least partially from re-entraining the circalunar oscillator).We thus compared the spawning rhythms of l-cry+/+ and l-cry–/– worms under a re-entrainment paradigm, where we provided our bright artificial culture full-moonlight at the time of the subjective new moon phase (Fig. 3a). In order to compare the spawning data distribution relative to the initial full moon (FM) stimulus, as well as to the new full moon stimulus (i.e. new FM), we used two nomenclatures for the months: months with numbers are analyzed relative to the initial nocturnal light stimulus (i.e. FM), whereas months with letters are analyzed relative to the new (phase-shifted) nocturnal light stimulus (i.e. new FM, Fig. 3a). When the nocturnal light stimulus is omitted (to test for the oscillator function) we then refer to ‘free-running FM’ (FR-FM) or ‘new free-running FM’ (new FR-FM), respectively (Fig. 3a). Using these definitions, the efficiency of circalunar clock re-entrainment will be reflected in the similarity of spawning data distributions between month 1 and month D, i.e. the more similar the distribution, the more the population has shifted to the new phase.Fig. 3: l-cry−/− mutants entrain the circalunar clock faster than wt to a high-intensity artificial moonlight stimulus.a Nocturnal moonlight exposure protocol of lunar phase shift (entrained by 8 nights, phased shifted by 6 nights of artificial culture moon, light green). b, c Number of mature animals (percent per month, rolling mean with a window of 3 days) of l-cry wild-type (b) and homozygous mutant (c) animals. p-values indicate results of Kolomogorov–Smirnov tests. Dark blue arrowheads- old FM phase: wt show a spawning minimum, indicative that the worms are not properly phase shifted. Mutants spawn in high numbers, but don’t spawn at the old NM indicated by light blue arrowhead. Also compare to initial FM and NM in months 1,2. d, e Circular plots of the data shown in (b) and (c). Each circle represents one lunar month. Each dot represents one mature worm. The arrow represents the mean vector characterized by the direction angle µ and r. r (length of µ) indicates phase coherence (measure of population synchrony). The inner circle represents the Rayleigh critical value (p = 0.05). f, g Results of two-sided multisample statistics of data in (d, e). Phase differences in days can be calculated from the angle between the two mean vectors (i.e. 12°= 1 day).Full size imageWhen using the artificial nocturnal light conditions, the re-entrainment of l-cry–/– animals was both faster and more complete than for their wildtype relatives, as predicted from our gate keeper hypothesis. This is evident from the linear data analysis and Kolmogorov–Smirnov tests when comparing the month before the entrainment (month 1) with two months that should be shifted after the entrainment (months C,D, Fig. 3b, c, f, g).Most notably, while l-cry−/− worms were fully shifted in month D (Fig. 3c: compare boxes and see complete lack of spawning at the light blue arrowhead indicating the old NM/new FR-FM phase versus massive spawning at new NM phase around dark blue arrowhead), wildtype animals were still mostly spawning according to the initial lunar phase (Fig. 3b: compare boxes and see spawning at the light blue arrowhead versus almost lack of spawning at dark blue arrowhead). The faster re-entrainment of l-cry–/–, compared to l-cry+/+ animals is also confirmed by the Mardia–Watson-Wheeler test (see Methods section for details). For l-cry+/+ animals, the comparisons of the spawning distributions before and after re-entrainment show a 1000-fold (months 1 versus C) and tenfold (months 1 versus D) higher statistical significance difference than the corresponding comparisons for l-cry−/− worms (Fig. 3f, g). Consistently, the phase differences in days calculated from the angle between the two mean vectors from the circular analysis is smaller in the mutants than in the wild types when comparing the phase of the month before the entrainment (month 1) with two months after the entrainment (months C, D) (Fig. 3d–g). The fact that there are still differences in the mutant population before and after entrainment is likely due to the fact that even the mutants are not fully re-entrained. However, they have shifted more robustly in response to an artificial nocturnal light stimulus than the wild types. This provides further evidence that in wildtype worms L-Cry indeed blocks the “wrong” light from entering into the circalunar clock and thus functions as a light gatekeeper.L-Cry functions mainly as light interpreter, while its contribution as direct moonlight entraining photoreceptor is (at best) minorWe next tested to which extent L-Cry is itself a sensor for the re-entrainment signal under naturalistic light conditions. Based on the finding that l-cry−/− worms can still re-entrain the circalunar oscillator (see above), it is clear that even if L-Cry also directly contributed to the entrainment, it cannot be the only moonlight receptor mediating entrainment. With the experiments below, we aimed to test if L-Cry has any role as an entraining photoreceptor to the monthly oscillator.Thus, we tested how the circalunar clock is shifted in response to a re-entrainment with naturalistic moonlight in Platynereis wt versus l-cry−/− worms. For this, animals initially raised and entrained under standard worm room light conditions of artificial sun- and moonlight (Supplementary Fig. 1b, e) were challenged by a deviating FM stimulus of 8 nights of naturalistic moonlight (Fig. 4a, Supplementary Fig. 1c, e). This re-entraining stimulus was repeated for three consecutive months (Fig. 4a).Fig. 4: l-cry has a minor contribution as entraining photoreceptor to circalunar clock entrainment.a Nocturnal moonlight exposure protocol of lunar phase shift with 8 nights of naturalistic moonlight (dark green). Number of mature animals (percent per month, rolling mean with a window of 3 days) of l-cry wild-type (b) and mutant (c) animals. p-values: Kolomogorov–Smirnov tests. Black arrowheads indicate spawning-free intervals of the wildtype, which shifted to the position of the new FM (under free-running conditions: FR-FM). d, e Data as in (b, c) plotted as circular data. 360° correspond to 30 days of the lunar month. The arrow represents the mean vector characterized by the direction angle µ and r. r (length of µ) indicates phase coherence (measure of population synchrony). p values are results of Rayleigh Tests: Significance indicates non-random distribution of data points. The inner circle represents the Rayleigh critical value (p = 0.05). f, g Results of two-sided multisample statistics on spawning data shown in (a–e). Phase differences in days can be calculated from the angle between the two mean vectors (i.e. 12°= 1 day).Full size imageThe resulting spawning distribution was analyzed for the efficacy of the naturalistic moonlight to phase-shift the circalunar oscillator. In order to test if the animals had shifted their spawning to the new phase, we again compared the spawning pattern before the exposure to the new full moon stimulus (months with numbers: data distribution analyzed relative to the initial/old FM, see Fig. 4a for an overview) to the spawning pattern after the exposure to the new full moon stimulus (months with letters: data distribution analyzed relative to the new FM, Fig. 4a). The more similar the data distributions of month 1 is to the months C, D, the more the population was shifted to the new phase.The first re-entraining full moon stimulus (Fig. 4b, first dark green box) is given in the middle of the main spawning period. The nocturnal light itself does not cause immediate effects on the number of spawning worms (Fig. 4b, see also Fig. 2b, c), but the repeated exposure resulted in a noticeable shift of the spawning distribution indicating a phase shift of the monthly oscillator in wildtype. Already at the third re-entraining full moon stimulus, wildtype animals exhibited a completely shifted spawning pattern (Fig. 4b, d-d″, month 1, 2 vs. month C). This is supported by statistical analyses: When comparing the months 1 and 2 (relative to the old FM before the shift) to the month C (relative to the new FM after the shift), both the Kolmogorov–Smirnov test (Fig. 4b: gray rectangles, 4f) and the Mardia–Watson–Wheeler test of the same data were non-significant (Fig. 4f), indicative of the population shifting to the new phase. Consistently, the direction angle (µ) of the mean vectors before and after the shift was highly similar, resulting in a phase difference of only 0.2 days between months 1 and C and 0.5 days between month 2 and month C (Fig. 4f, for details see methods). The month under circalunar free-running conditions (month D) supports this observation, albeit with lower statistical support (Fig. 4b, d″, f).Of note, wild-type worms would eventually reach the high spawning precision found under naturalistic moonlight only after several more months based on independent experiments (Fig. 2j, k).When we analyzed the spawning distribution of l-cry mutants in the same way as the wild types, we found that the data distribution exhibited significant differences in the linear Kolmogorov–Smirnov test when comparing months 1 and 2 before the shift to the months C and D after the shift (Fig. 4c: gray rectangles, Fig. 4g); as well as in the phase distribution in the circular analyses when comparing the months before the shift (months 1 and 2) with the last months of the shift (months C,D) (Fig. 4e, e′ versus e″, e‴, g). The populations also exhibited a noticeable phase difference of ≥3.5 days (Fig. 4g).Based on the statistical significant difference in the re-entrainment of l-cry–/–, but not wild-type populations under a naturalistic sun- and moonlight regime, we conclude that L-Cry also likely contributes to circalunar entrainment as a photoreceptor. However, as these differences are rather minor, compared to the much stronger differences seen under artificial light regime, we conclude that its major role is the light gatekeeping function.In an independent study that focused on the impact of moonlight on daily timing, we identified r-Opsin1 as a lunar light receptor that mediates moonlight effects on the worms’ ~24 h clock36. We tested if r-opsin1 is similarly important for mediating the moonlight effects on the monthly oscillator of the worm, analyzed here. This is not the case. r-opsin1–/– animals re-entrain as well as wildtype worms under naturalistic light conditions (Supplementary Fig. 4). This adds to and is also consistent with our above observation that the spawning distribution is un-altered between r-opsin1–/– and wildtype animals under artificial light conditions (Supplementary Fig. 3a, b, e, f). This finding also further enforces the notion that monthly and daily oscillators use distinct mechanisms, but both require L-Cry as light interpreter.L-Cry discriminates between naturalistic sun- and moonlight by forming differently photoreduced statesGiven that the phenotype of l-cry–/– animals suggests a role of L-Cry as light gatekeeper, i.e. only allowing the ‘right’ light to most efficiently impact on the circalunar oscillator, we next investigated how this could function on the biochemical and cell biological level.While we have previously shown that Pdu-L-Cry is degraded upon light exposure in S2 cell culture15, it has remained unclear if L-Cry has the spectral properties and sensitivity to sense moonlight and whether this would differ from sunlight sensation. To test this, we purified full length L-Cry from insect cells (Supplementary Fig. 5a–c). Multi-angle light scattering (SEC-MALS) analyses of purified dark-state L-Cry revealed a molar mass of about 130 kDa, consistent with the formation of an L-Cry homodimer (theoretical molar mass of L-Cry monomer is 65.6 kDa) (Fig. 5a). Furthermore, purified L-Cry binds Flavin Adenine Dinucleotide (FAD) as its chromophore (Supplementary Fig. 5d, e). We then used UV/Vis absorption spectroscopy to analyze the FAD photoreaction of purified L-Cry in presence of 1 mM TCEP to prevent protein oxidation. The absorption spectrum of dark-state L-Cry showed maxima at 450 nm and 475 nm, consistent with the presence of oxidized FAD (Supplementary Fig. 5f, black line). As basic starting point to analyze its photocycle, L-Cry was photoreduced using a LED (PerkinElmer ACULED Dyo) with a blue-light dominated spectrum and spectral peak at 450 nm (Supplementary Fig. 1d, d′, henceforth referred to as “blue-light”) for 110 s37. The light-activated spectrum showed that blue-light irradiation of L-Cry leads to the complete conversion of FADox into an anionic FAD radical (FADo-) with characteristic FADo- absorption maxima at 370 nm and 404 nm and reduced absorbance at 450 nm (Supplementary Fig. 5f, blue spectrum, black arrows). In darkness, L-Cry reverted back to the dark-state with time constants of 2 min (18 °C), 4 min (6 °C) and 4.7 min (ice) (Supplementary Fig. 5g–k).Fig. 5: L-Cry forms differently photoreduced sunlight- and moonlight states.a Multi-Angle Light Scattering (MALS) analyses of dark-state L-Cry fractionated by size exclusion chromatography (SEC). Black dashed line: normalized UV absorbance, solid line: normalized scattering signal. The molar mass of about 130 kDa derived from MALS (mass signal shown in red) corresponds to an L-Cry homodimer. b Absorption spectrum of L-Cry in darkness (black) and after sunlight exposure (orange). Additional timepoints: Supplementary Fig. 6a. c Dark recovery of L-Cry after 20 min sunlight on ice. Absorbance at 450 nm in Supplementary Fig. 6b. d, e Absorption spectra of L-Cry after exposure to naturalistic moonlight for different durations. f Full spectra of dark recovery after 6 h moonlight. Absorbance at 450 nm: Supplementary Fig. 6d. g Absorption spectrum of L-Cry after 6 h of moonlight followed by 20 min of sunlight. h Absorption spectrum of L-Cry after 20 min sunlight followed by moonlight first results in dark-state recovery. Absorbance at 450 nm: Supplementary Fig. 6e. i Absorption spectrum of L-Cry after 20 min sunlight followed by 4 h and 6 h moonlight builds up the moonlight state. j Model of L-Cry responses to sunlight (orange), moonlight (green) and darkness (black). Only transitions between stably accumulating states are shown. Absorbances in (b–i) were normalized when a shift in the baseline occurred between different measurements of the same measurement set, which is then indicated on the Y-axis as “normalized absorbance”.Full size imageWe then investigated the response of L-Cry to ecologically relevant light, i.e. sun- and moonlight using naturalistic sun- and moonlight devices that we designed based on light measurements at the natural habitat of P. dumerilii31 (Supplementary Fig. 1a, c, e). Upon naturalistic sunlight illumination, FAD was photoreduced to FADo-, but with slower kinetics than under the stronger blue-light source, likely due to the intensity differences between the two lights (Supplementary Fig. 1c–e).While blue-light illumination led to a complete photoreduction within 110 s (Supplementary Fig. 5f), sunlight induced photoreduction to FADo- was completed after 5–20 min (Fig. 5b) and did not further increase upon continued illumation for up to 2 h (Supplementary Fig. 6a). Dark recovery kinetics had time constants of 3.2 min (18 °C) and 5 min (ice) (Fig. 1c, Supplementary Fig. 6b, c).As the absorbance spectrum of L-Cry overlaps with that of moonlight at the Platynereis natural habitat (Supplementary Fig. 1a), L-Cry has the principle spectral prerequisite to sense moonlight. However, the most striking characteristic of moonlight is its very low intensity (5.8 × 1010 photons/cm2/s at −5m, Supplementary Fig. 1a–e). To test if Pdu-L-Cry is sensitive enough for moonlight, we illuminated purified L-Cry with our custom-built naturalistic moonlight, closely resembling full-moonlight intensity and spectrum at the Platynereis natural habitat (Supplementary Fig. 1a, c, e). Naturalistic moonlight exposure up to 2.75 h did not markedly photoreduce FAD, notably there was no difference between 1 h and 2.75 h (Fig. 5d). However, further continuous naturalistic moonlight illumination of 4 h and longer resulted in significant changes (Fig. 5d), whereby the spectrum transitioned towards the light activated state of FADo- (note peak changes at 404 nm and at 450 nm). This photoreduction progressed further until 6 h naturalistic moonlight exposure (Fig. 5d). No additional photoreduction could be observed after 9 h and 12 h of naturalistic moonlight exposure (Fig. 5e), indicating a distinct state induced by naturalistic moonlight that reaches its maximum after ~6 h, when about half of the L-Cry molecules are photoreduced. This time of ~6 h is remarkably consistent with classical work showing that a minimum of ~6 h of continuous nocturnal light is important for circalunar clock entrainment, irrespective of the preceding photoperiod5. The dark recovery of L-Cry after 6 h moonlight exposure occurred with a time constant of 6.7 min at 18 °C (Fig. 5f, Supplementary Fig. 6d). Given that both sunlight and moonlight cause FAD photoreduction, but with different kinetics and different final FADo- product/FADox educt ratios, we wondered how purified L-Cry would react to transitions between naturalistic sun- and moonlight (i.e. during “sunrise” and “sunset”).Mimicking the sunrise scenario, L-Cry was first illuminated with naturalistic moonlight for 6 h followed by 20 min of sunlight exposure. This resulted in an immediate enrichment of the FADo- state (Fig. 5g). Hence, naturalistic sunlight immediately photoreduces remaining oxidized flavin molecules, that are characteristic of moonlight activated L-Cry, to FADo-, to reach a distinct fully reduced sunlight state.In contrast, when we next mimicked the day-night transition (“sunset”) by first photoreducing with naturalistic sunlight (or strong blue-light) and subsequently exposed L-Cry to moonlight, L-Cry first returned to its full dark-state within about 30 min (naturalistic sunlight: τ = 7 min (ice), Fig. 5h, Supplementary Fig. 6e; blue-light: τ = 9 min (ice), Supplementary Fig. 6f–h), despite the continuous naturalistic moonlight illumination. Prolonged moonlight illumination then led to the conversion of dark-state L-Cry to the moonlight state (Fig. 5i, Supplementary Fig. 6f). Hence, fully photoreduced sunlight-state L-Cry first has to return to the dark-state before accumulating the moonlight state characterized by the stable presence of the partial FADo- product/FADox educt. In contrast to sunlight-state L-Cry, moonlight-state L-Cry does not return to the oxidized (dark) state under naturalistic moonlight (Fig. 5e), i.e. moonlight maintains the moonlight state, but not the sunlight state. We note, that a partially photoreduced L-Cry state may be formed transiently during dark-state recovery of the sunlight state under moonlight. However, this transiently occurring partially photoreduced L-Cry state would differ from the “true” moonlight state (e.g. by an allosteric change) preventing its accumulation (see discussion and Supplementary Fig. 6i).Given that L-Cry forms a homodimer and moonlight photoreduces about half of the FAD molecules, we propose that the moonlight state corresponds to a half-reduced FADo- FADox dimer, where FAD is only photoreduced in one L-Cry monomer, whereas in the sunlight state both monomers are photoreduced (FADo- FADo-) (Fig. 5j). This implies that the quantum yield for FADox to FADo- photoreduction differs between the two L-Cry monomers. One monomer (referred to as “A” in Fig. 5j) acts as “very low intensity light sensor” with a high quantum yield ΦA. Hence, the very low photon number provided after 6 h of moonlight illumination is sufficient to photoreduce its flavin co-factor, resulting in the partially photoreduced FADo- FADox moonlight state (Fig. 5j).For direct comparison, our naturalistic moonlight’s emission (in the main absorbance range of L-Cry: 330 nm–510 nm) is 5.4 × 1010 photons/cm2/s (Supplementary Fig. 1e), which accumulates to ~1.2 × 1015 photons/cm2 in the 6 h required to reach the half-reduced moonlight state (Fig. 5d, e). For naturalistic sunlight, emitting ~7.5 × 1014 photons/cm2/s (330–510 nm), at least 5 min of sunlight illumination (i.e. > ~1.8 × 1017 photons/cm2) are required to photoreduce the flavin in both L-Cry monomers in order to reach the fully photoreduced FADo- FADo- sunlight state (Fig. 5b, j). Thus, the second L-Cry monomer (monomer “B” in Fig. 5j) has a significantly lower quantum yield ΦB for FAD photoreduction (ΦB More
