Cryptochromes integrate green light signals into the circadian system

Abstract Plants are acutely sensitive of their light environment, adapting their growth habit and prioritizing developmental decisions to maximize fecundity. In addition to providing an energy source and directional information, light quality also contributes to entrainment of the circadian system, an endogenous timing mechanism that integrates endogenous and environmental signalling cues to promote growth. Whereas plants' perception of red and blue portions of the spectrum are well defined, green light sensitivity remains enigmatic. In this study, we show that low fluence rates of green light are sufficient to entrain and maintain circadian rhythms in Arabidopsis and that cryptochromes contribute to this response. Importantly, green light responses are distinguishable from low blue light‐induced phenotypes. These data suggest a distinct signalling mechanism enables entrainment of the circadian system in green light‐enriched environments, such as those found in undergrowth and in densely planted monoculture.


| INTRODUCTION
Plants are highly sensitive to changes in ambient light conditions, with a complex photosensory network evolving to facilitate the cellautonomous perception of light across the electromagnetic spectrum.
In addition to being used as an energy source, plants decipher the composition and duration of light perceived to make appropriate developmental decisions. Whereas photomorphogenesis is a critical component of early development, light also informs development in mature tissues (Whitelam & Halliday, 2007). In general, red and blue light enable plants to orientate themselves appropriately within a canopy, whereas far-red and green-enriched light is perceived as an indication of overgrowing vegetation, inducing a shade avoidance response (Casal, 2013;Liscum et al., 2014;Wang, Zhang, & Folta, 2015;Zhang & Folta, 2012). In combination, these responses allow plants to optimize their light-gathering capacity.
Prior plant biology literature has subdivided the light spectrum into ultraviolet (UV) (320-400 nm), blue (400-500 nm), green (500-600 nm), red (600-700 nm), and far-red portions (700-800 nm). Although UV, blue, and red/far-red photoreceptors have been identified and well characterized, specific green photoreceptors have yet to be identified in higher plants (Christie, Blackwood, Petersen, & Sullivan, 2015;Rizzini et al., 2011;. Instead, our current understanding suggests that plants perceive green light through the residual sensitivity of blue and red photoreceptors, with phytochromes and cryptochromes absorbing portions of the green spectrum, albeit at a fraction of the sensitivity towards their primary wavelength (Sellaro Mageroy, Justice, & Folta, 2013). In addition, plants are able to harvest green light for photosynthesis via both chlorophylls and carotenoids (Smith et al., 2017), whereas a distinct role for zeaxanthin as a blue/green reversible photoreceptive pigment in guard cells has also been proposed (Frechilla, Talbott, Bogomolni, & Zeiger, 2000;Talbott et al., 2006;Talbott, Zhu, Han, & Zeiger, 2002). Seedlings' perception of blue:green ratios regulates hypocotyl extension, suggesting that plants interpret blue:green ratios as a shade response, whereas green light has also been reported to antagonize blue-and UV-B induced stomatal opening (Casal, 2012;Eisinger, Bogomolni, & Taiz, 2003;Sellaro et al., 2010;Smith et al., 2017;Talbott et al., 2006). These reports highlight the importance of understanding how plants respond to complex, multichromatic lighting regimes.
Cryptochromes have previously been reported to perceive both blue and green portions of the spectrum (Ahmad et al., 2002;Lin, Ahmad, Gordon, & Cashmore, 1995). Although dark-adapted cryptochromes do not absorb light wavelengths longer than 500 nm, illuminated cryptochrome photocycle intermediates absorb light up to 650 nm (Banerjee et al., 2007;Bouly et al., 2007). It has subsequently been proposed that shorter wavelengths of green light (<530 nm) are perceived as part of the canonical cryptochrome and phototropin blue light response, whereas longer wavelengths of green/yellow light (~570 nm) accelerate the reversion of blue-light activated cryptochrome to its inactive state (Bouly et al., 2007). This latter hypothesis provides an elegant photochemical explanation for the observed antagonization of blue photoperception by longer wavelengths of "green" light, although the photochemical mechanism underlying this remains elusive (Banerjee et al., 2007;Bouly et al., 2007;Herbel et al., 2013;. In addition to its role in development and photosynthesis, light quality and intensity also informs progression of the circadian system, an endogenous timing mechanism that coordinates metabolism, physiology, and development with prevailing environmental conditions. The pace, phase, and amplitude of the central oscillator are regulated by the quality and intensity of light irradiation, with both photoreceptors and photoassimilates contributing to the maintenance of circadian rhythms (Baudry et al., 2010;Haydon, Mielczarek, Robertson, Hubbard, & Webb, 2013;Somers, Devlin, & Kay, 1998). Genes such as CCA1 and PRR9 are induced by light, whereas PRR7 and GIGANTEA monitor photoassimilate accumulation and modulate circadian timing accordingly (Haydon et al., 2013;Haydon, Mielczarek, Frank, Román, & Webb, 2017;Ito et al., 2003;Locke et al., 2005;Wang & Tobin, 1998). As a consequence, the circadian system provides a well understood readout of plant photoperception in addition to its role as a governor of plant development.
Here, we examine the role of cryptochromes as blue/green photoreceptors within the circadian system. We demonstrate that both cry1 and cry2 contribute to the clock's response to green light. Our data also distinguish between blue and green light-induced phenotypes, suggesting that green light sensitivity is not merely a consequence of residual chromophore absorption above 500 nm. Finally, we determine that cry mutants continue to have a phenotype under blue/green light, suggesting shorter wavelengths of green light are insufficient to impair cry-dependent blue light signalling into the circadian system.

| Hypocotyl measurements
Seeds were irradiated with cool fluorescent white light at 60 μmol m −2 s −1 for 4 hr before being moved to coloured LEDs as per experimental requirements and grown vertically for 5 days before being imaged and processed using ImageJ (Schneider, Rasband, & Eliceiri, 2012). The length of hypocotyls was normalized to the average length of a darkgrown control.

| Luciferase imaging
Plants were entrained for 6 days in 12:12 L/D cycles under white light on MS medium without sucrose before being sprayed with 3-mM Dluciferin in 0.01% (v/v) Triton X-100 as previously described (Litthauer, Battle, Lawson, & Jones, 2015). Experiments performed using CO 2 -depleted air were completed as previously described (Kircher & Schopfer, 2012). In brief, 5-g sodalime was added to a double-sealed bag enclosing the petri plate on which seedlings had been sown immediately before circadian imaging. Imaging was completed over 5 days using an Andor iKon-M CCD camera controlled by μManager (Edelstein, Amodaj, Hoover, Vale, & Stuurman, 2010) before data were processed using ImageJ (Schneider et al., 2012).
Patterns of luciferase activity were fitted to cosine waves using fast

| Real-time reverse transcription polymerase chain reaction
Following entrainment, plants were transferred to 20 μmol m −2 s −1 blue light or green light provided by LEDs. Tissue was harvested at the indicated time before RNA was isolated from 10 to 15 seedlings for each data point using Tri Reagent® according to the manufacturer's protocol (Sigma Aldrich, Dorset, UK, http://www. sigmaaldrich. com). Reverse transcription was performed using RevertAid reverse transcriptase following DNase treatment (Fisher Scientific, Loughborough, UK, http://www.fisher.co.uk). Real-time reverse transcription polymerase chain reaction was performed using a BioRad CFX96 Real-Time system. Samples were run in triplicate, with starting quantity estimated from critical thresholds using the standard curve of amplification. Data for each sample were normalized to APX3, IPP2, and At1g11910 expression as internal controls as previously described (Nusinow et al., 2011). Primer sets used are described in Table S1.

| Green light maintains circadian rhythmicity
Plants sense light via specific photoreceptors and also indirectly via the acquisition of photoassimilates from photosynthesis (Jones, 2017). The red light-sensitive phytochromes, as well as the blue light-sensitive cryptochrome and phototropin families have been well described (Christie et al., 2015;Rockwell, Su, & Lagarias, 2006), but previous studies have described only limited roles for green light in plant development . In our study, we used a combination of green LEDs and cut-off filters to illuminate plants with broadband green light (500-600 nm, Figure 1a). As previously Although we did not observe significant differences in hypocotyl elongation in response to increasing green light, we were curious whether green light was sufficient to maintain circadian rhythms of gene expression. Transgenic Arabidopsis seedlings expressing a luciferase circadian reporter maintain circadian rhythms for multiple days when transferred to constant white, blue, or red light (Millar, Straume, Chory, Chua, & Kay, 1995). By contrast, in the absence of light, or under dim blue light (1 μmol m −2 s −1 ), circadian rhythms of luciferase bioluminescence dampened to apparent arrhythmia within 24 hr in the absence of sucrose (Haydon et al., 2017 Figure S1). To evaluate the role of green light in circadian rhythms, plants carrying a bioluminescent circadian reporter (CCA1:: LUC2 or TOC1::LUC ) were entrained to 12:12 light:dark cycles before being transferred to 10 μmol m −2 s −1 constant green light (Figure 1d,e). As for blue and red light, we observed that green light was sufficient to sustain circadian rhythms of luciferase activity with both reporters, with τ = 28.17 ± 0.26 hr with CCA1:: LUC2 or τ = 27.65 ± 0.40 hr in TOC1::LUC seedlings, respectively. Such data demonstrate that dim green light is sufficient to maintain circadian rhythms, despite not inhibiting hypocotyl elongation.

| The circadian system is responsive to green light
In order to better understand the effect of green light upon the circadian system, we completed a fluence rate response curve (  Figure S2). We also examined whether green light was sufficient to entrain the circadian system ( Figure 2d). Plants were entrained under white light before being transferred to alternating periods of 12 hr green light, 12 hr darkness. Following 24 hr under these conditions, dawn was delayed by 12 hr so that plants experienced an extended night. Plants treated in this way were able to entrain to the revised timing of dawn (Figures 2d and S2d), demonstrating that the circadian system is responsive to green light, either via a green photoreceptor or as a consequence of green light-derived photosynthesis.

| Photoactivated cryptochromes contribute to green light signalling into the circadian system
Of the known photoreceptors, cryptochromes have previously been described as blue/green photoreceptors, whereas phytochromes are also activated by green light Shinomura et al., 1996). Although phyA-211 and phyB-9 mutants did not have a circadian phenotype under constant green light ( Figure S3), we observed a significant extension of circadian period in cryptochrome mutants under these conditions (p < .01, Dunnett's test, Figure 3a A comparison of absorbance spectra from dark-adapted or illuminated cryptochromes indicate that these photoreceptors absorb proportionally more green light following illumination (Banerjee et al., 2007;Bouly et al., 2007). We were subsequently curious if cry seedlings transferred immediately to green light from the dark retained a circadian phenotype. The half-life of photoactivated cryptochromes has been estimated to be approximately 6 min (Herbel et al., 2013), and so we

| Exogenous sucrose is sufficient to rescue the cryptochrome circadian phenotype under green light
Although our data suggest a role for cryptochromes in green light perception, recent work has emphasized the contribution of photoassimilates to circadian timing (Frank et al., 2018;Haydon et al., 2013;Haydon et al., 2017;Philippou, Ronald, Sanchez-Villarreal, Davis, & Davis, 2019). In order to assess the contribution of photosynthesis towards circadian rhythmicity under constant green light, we assessed circadian rhythms in the presence of CO 2 -depleted air, and/or by supplying exogenous sucrose within the growth media to saturate the cellular response to photosynthetically derived sucrose (Figures 2d and 4). As under constant dichromatic blue and red light (Haydon et al., 2013)    shade has begun to emerge in recent years .

| Cryptochromes continue to signal into the circadian system in the presence of blue/green light
Phytochromes, phototropins, and cryptochromes have each been implicated in specific green light responses ranging from hypocotyl growth inhibition to petiole elongation (Zhang, Maruhnich, & Folta, 2011). Our work with short wavelength green light reveals that green light is additionally sufficient to maintain circadian rhythms, despite this portion of the electromagnetic spectrum having little effect upon the inhibition of hypocotyl elongation (Figures 1 and 2).
Cryptochromes were originally identified as blue/green photoreceptors, although the effect of green light upon cryptochrome photoperception has proven to be complex. Plants overexpressing CRY1 have previously been reported as being hypersensitive to short-wavelength green light (<532 nm, Ahmad et al., 2002, Bouly et al., 2007 whereas long-wavelength green light centered around 570 nm is sufficient to antagonize cryptochrome activation (Banerjee et al., 2007;Bouly et al., 2007;Herbel et al., 2013). Our experiments, utilizing a short-wavelength green light (peak 527 nm), support the hypothesis that cryptochrome signalling is activated in the presence of this portion of the spectra whereas phyA and phyB seedlings did not have a circadian phenotype ( Figure S3). It has previously been proposed that cryptochrome green light sensitivity either arises from residual sensitivity of the bound flavin chromophore at wavelengths longer than 500 nm (Ahmad et al., 2002;, or that irradiated cryptochromes absorb green light as part of their photocycle (Banerjee et al., 2007;Bouly et al., 2007). Our work comparing dark-and light-adapted seedlings (Figure 3) suggests that irradiated cryptochromes contribute to the integration of green light signals into the circadian system, although the photochemistry underlying this phenotype remains to be investigated.

| Exogenous sucrose masks the contribution of cryptochromes to circadian FRP under constant green light
Interpretation of light signalling into the circadian system is complicated by the clock's response to metabolites derived from photosynthesis (Frank et al., 2018;Haydon et al., 2013;Haydon et al., 2017;Philippou et al., 2019). In order to assess the contribution of photosynthesis towards circadian rhythmicity under constant green light, we assessed circadian rhythms in the presence of CO 2 -depleted air (Figures 2d and 4). cry1 seedlings retained an extended circadian FRP in CO 2 -depleted conditions, although cry2 seedlings were indistinguishable from wild type plants in a reduced CO 2 environment.
Such data suggests that cry1 acts in parallel to photosynthate-derived signals to regulate circadian FRP under constant green light. Exogenous sucrose has previously been used to saturate plants' circadian responses to photoassimilates (Frank et al., 2018;Haydon et al., 2013;Haydon et al., 2017;Philippou et al., 2019). Interestingly, exogenous sucrose was sufficient to mask the cry1 circadian defect

| Cryptochrome signalling into the circadian system is distinct under either blue light and green light
Although cryptochromes contribute to both blue-and green-light signalling pathways into the circadian system (Figures 3, 4, and 5 (Belbin et al., 2016) but differ from earlier reports that included sucrose as a media additive . By contrast, loss of cryptochrome function under green light causes an extension of circadian period without an associated loss of amplitude that is masked by the addition of supplemental sucrose (Figures 3, 4, 5, and S6b).
These data suggest either that cryptochromes have distinct roles in circadian responses to blue or green light or that additional photoreceptors, such as phytochromes, additively contribute to circadian perception of green light.
Although it is difficult to directly compare different lighting regimes, we were interested to note that a combination of blue and green light was able to maintain the amplitude of bioluminescence in cry1 seedlings with an extended circadian FRP phenotype (Figure 6 c-e). Such data emphasize the ability of plants to perceive and integrate information from across the light spectrum to respond to prevailing environmental conditions. Previous work has identified physical interactions between cryptochromes and phytochromes, as well as between the signalling cascades induced by these photoreceptors (Hughes, Vrana, Song, & Tucker, 2012;Mas, Devlin, Panda, & Kay, 2000;Pedmale et al., 2016;Wang et al., 2018). It will consequently be of great interest to determine how phytochromes and cryptochromes interact to appropriately respond to green light as part of plants' complex response to natural illumination.