Cryptochrome-mediated blue-light signalling modulates UVR8 photoreceptor activity and contributes to UV-B tolerance in Arabidopsis

UV-B constitutes a critical part of the sunlight reaching the earth surface. The homodimeric plant UV-B photoreceptor UV RESISTANCE LOCUS 8 (UVR8) monomerizes in response to UV-B and induces photomorphogenic responses, including UV-B acclimation and tolerance. REPRESSOR OF UV-B PHOTOMORPHOGENESIS 1 (RUP1) and RUP2 are negative feedback regulators that operate by facilitating UVR8 ground state reversion through re-dimerization. Here we show that RUP1 and RUP2 are transcriptionally induced by cryptochrome photoreceptors in response to blue light, which is dependent on the bZIP transcriptional regulator ELONGATED HYPOCOTYL 5 (HY5). Elevated RUP1 and RUP2 levels under blue light enhance UVR8 re-dimerization, thereby negatively regulating UVR8 signalling and providing photoreceptor pathway cross-regulation in a polychromatic light environment, as is the case in nature. We further show that cryptochrome 1, as well as the red-light photoreceptor phytochrome B, contribute to UV-B tolerance redundantly with UVR8. Thus, photoreceptors for both visible light and UV-B regulate UV-B tolerance through an intricate interplay allowing the integration of diverse sunlight signals.

S unlight is an essential factor for life on earth. It is the ultimate source of energy for photosynthesis and it modulates plant growth and development. Flowering plants can perceive light signals spanning from UV-B to far-red (~280-800 nm) using five classes of photoreceptors 1 . UV-B (280-315 nm) induces photomorphogenic responses and UV-B acclimation [2][3][4][5] . The UV-B photoreceptor UV RESISTANCE LOCUS 8 (UVR8) forms homodimers in its ground state and monomerizes to the active form in response to UV-B absorption 6,7 . UVR8 monomers interact directly with the WD40 domain of the E3 ubiquitin ligase CONSTITUTIVELY PHOTOMORPHOGENIC 1 (COP1) 5,7 . The UVR8−COP1 interaction competes with that between COP1 and the photomorphogenesis-promoting bZIP transcription factor ELONGATED HYPOCOTYL 5 (HY5), thereby preventing its ubiquitination 8,9 . This leads to stabilization of HY5 and the induction of many target genes encoding proteins involved in the photomorphogenic response to UV-B, including HY5 itself [10][11][12] . Among these HY5-induced genes, REPRESSOR OF UV-B PHO-TOMORPHOGENESIS 1 (RUP1) and RUP2 encode WD40-repeat proteins that provide UVR8 negative feedback regulation 13 . RUP1 and RUP2 directly interact with UVR8 to facilitate its redimerization, thereby inactivating the UVR8 monomer 14,15 . RUP1 and RUP2 can also be part of a CUL4-DDB1-based E3 ubiquitin ligase that targets HY5 for degradation 16 . Moreover, it has been proposed that COP1 directly targets RUP1 and RUP2 for ubiquitination and degradation under UV-B, contributing to the stabilization of HY5 16 .
Here, we show that induction of RUP1 and RUP2 gene expression and their ensuing protein accumulation are blue-light responsive. These inductions depend mainly on the blue-light photoreceptor cry1, through the activity of HY5, with lesser roles played by cry2 and phyA. Enhanced RUP1 and RUP2 levels under blue light affect the balance between UVR8 monomer and UVR8 homodimer, thereby modulating the activity of the UV-B signalling pathway. Finally, we demonstrate that cry1, phyB, and UVR8 redundantly regulate UV-B tolerance.

Results
Cryptochromes and phyA activate RUP1 and RUP2 expression. Blue-light exposure of Arabidopsis seedlings resulted in strong and transient induction of RUP1 and RUP2 expression in wild type, but not in hy5 (Fig. 1a, b). In agreement, RUP2 protein accumulated in response to blue light in wild type, but not in hy5 to a detectable level (Fig. 1c). To identify the photoreceptors responsible for the blue-light induction of RUP1 and RUP2 expression, we examined responses in cry1, cry2, cry1 cry2, and cry1 cry2 phyA. cry1 and cry2 single mutants both displayed reduced blue-light induction of RUP1 and RUP2, which was further reduced in cry1 cry2 double mutants and absent in cry1 cry2 phyA triple mutants (Fig. 1d, e). In agreement, RUP2 protein accumulation in response to blue light was reduced in cry1, cry2, and phyA, further reduced in cry1 cry2, and undetectable in cry1 cry2 phyA (Fig. 1f). The absence of an anti-RUP1 antibody prevented directly testing endogenous RUP1 levels. We conclude that blue-light-dependent cryptochrome and phyA signalling activates RUP1 and RUP2 expression, resulting in RUP2, and likely RUP1, protein accumulation. cry1 and phyA signalling enhances UVR8 re-dimerization. Accumulation of RUP1 and RUP2 in response to blue light points to a previously unknown effect of blue-light signalling on UVR8 activity. We thus tested the effect of blue light on the dynamics of the UVR8 homodimer/monomer ratio upon UV-B treatment, with a particular focus on UVR8 re-dimerization post UV-B exposure (Fig. 2a). The UV-B treatment induced a strong UVR8 monomerization in wild type and rup1 rup2 but did not affect the total amount of UVR8 (−UV and +UV; Fig. 2b, c). During the subsequent recovery in darkness, UVR8 re-dimerization was significantly faster in wild-type seedlings that were pre-exposed to blue light than that in seedlings without blue-light treatment (30′ and 60′; Fig. 2b, c). This blue-light enhancement of UVR8 redimerization was absent in rup1 rup2 mutants, in which UVR8 remained monomeric during the 1-h post-UV-B recovery (Fig. 2b, c). UVR8 re-dimerization was also impaired in rup1 and rup2 single mutants but to a considerably lesser extent than in the rup1 rup2 double mutant, supporting redundant roles of RUP1 and RUP2 in this process (Supplementary Fig. 1) 13,14 . Moreover, in agreement with the role of HY5 in RUP1 and RUP2 activation, blue-light-enhanced UVR8 re-dimerization was strongly impaired in hy5 mutants ( Supplementary Fig. 2).
To test the roles of specific photoreceptors in blue-lightenhanced UVR8 re-dimerization, we quantified the UVR8 dimer / UVR8 total ratio in wild-type seedlings and cry1, cry2, cry1 cry2, phyA, and cry1 cry2 phyA mutant seedlings. Without blue-light pre-treatment, no differences were observed in UVR8 redimerization kinetics among the six tested genotypes (Fig. 2d, f). However, in blue-light pre-treated seedlings, the enhanced UVR8 re-dimerization apparent in wild type was partially impaired in phyA and considerably more so in cry1 cry2 and cry1 cry2 phyA (Fig. 2e). This result agrees with the reduced RUP2 accumulation in response to blue light observed in these photoreceptor mutant lines (Fig. 1e). In further agreement with a key role of cry1, overexpression of GFP-cry1 resulted in elevated levels of RUP2 in response to blue light and further enhanced UVR8 re-dimerization when compared to wild type under these conditions ( Supplementary Fig. 3). In contrast to cry1 and cry1 cry2, blue-light pre-treated cry2 single mutants displayed UVR8 re-dimerization kinetics comparable to that in wild type (Fig. 2g). We conclude that blue light enhances UVR8 re-dimerization, and that this is mainly due to cry1 and phyA signalling-mediated RUP1 and RUP2 accumulation.
BIC overexpression blocks enhanced UVR8 re-dimerization. BIC1 and BIC2 are negative feedback regulators of cry1 and cry2 17 . We therefore tested whether BIC1 and BIC2 accumulation could affect RUP1/RUP2 expression and RUP2 protein levels. Blue-light-induced accumulation of RUP1/RUP2 mRNA and RUP2 protein was indeed reduced in the BIC1 and BIC2 overexpression lines bic1-D1, BIC1-GFP, and BIC2-GFP to a similar level as in cry1 cry2 (Fig. 3a, b). This agrees with inhibitory roles of BIC1 and BIC2 in cry1 and cry2 activities 17 . However, we did not observe enhanced RUP2 protein levels in the bic1 bic2 double mutant compared to that in wild type (Fig. 3b). To test whether BIC1 and BIC2 overexpression modulates UVR8 activity, we UV-B treated various genotypes that were pre-treated with blue light or not, and then quantified UVR8 re-dimerization after 30-min recovery. In agreement with the observed RUP2 levels, blue-light-enhanced UVR8 re-dimerization was similar in bic1 bic2 and wild type, but strongly impaired in the BIC1 and BIC2 overexpression lines, comparable to that in cry1 cry2 and rup1 rup2 (Fig. 3c). Interestingly, reciprocally to RUP1 and RUP2 activation through cryptochrome-mediated blue-light signalling ( Fig. 1) 13 , BIC1 and BIC2 are induced by UVR8-mediated UV-B signalling (Fig. 3d) 5 . We conclude that inhibition of cryptochrome activity by BIC1 and BIC2 suppresses the blue-light enhancement of UVR8 re-dimerization, and that UVR8-mediated regulation of BIC1 and BIC2 may add an additional regulatory control level. cry1 and cry2 repress UVR8-induced responses. The impact of blue-light signalling on UVR8 activity indicates an associated impact on UV-B responses. In comparison to wild type, cry1 cry2 double mutants showed enhanced hypocotyl growth inhibition under supplementary UV-B ( Supplementary Fig. 4a). It is of note, however, that cry1 cry2 shows an elongated hypocotyl in the absence of UV-B compared to wild type, and a similar hypocotyl length in the presence of UV-B ( Supplementary Fig. 4a). We thus reduced the UV-B to levels that were insufficient to induce hypocotyl growth inhibition in wild type to further test for UV-B hypersensitivity in cry1 cry2. Strikingly, cry1 cry2 displayed a clear hypocotyl growth inhibition response under these conditions (Fig. 4a). This was further confirmed to be the case for a cry1 mutant in the Ws accession ( Supplementary Fig. 4b). A minor difference was also observed in UV-B-responsive anthocyanin accumulation between wild type and cry1 cry2 under these conditions (Fig. 4b). We further investigated expression of UV-Bresponsive genes. We first compared the induction of HY5 and CHALCONE SYNTHASE (CHS) in cry1 cry2, uvr8, and wild type in response to 3-h supplemental narrowband UV-B 5  . c Immunoblot analysis of RUP2 protein level in 4-d-old Col and hy5 seedlings grown in darkness (0) or treated with blue light for 6 or 12 h. The asterisk indicates a nonspecific cross-reacting band. Actin is shown as protein loading control. d, e qRT-PCR analysis of d RUP1 and e RUP2 expression in 4-d-old Col, cry1-304 (cry1), cry2-1 (cry2), cry1-304 cry2-1 (cry1 cry2), and cry1-304 cry2-1 phyA-412 (cry1 cry2 phyA) seedlings grown in darkness or treated with blue light for 1, 3, 6, or 9 h. Data are means ± SEM (N = 3). Note that the Col samples in panels (d) and (e) are identical to those in panels (a) and (b), respectively, as the data are derived from the same experiment. f Immunoblot analysis of RUP2 protein level in 4-d-old Col, cry1, cry2, cry1 cry2, phyA-211 (phyA), cry1 cry2 phyA, and rup1 rup2 seedlings grown in darkness, then treated with blue light for 12 h (+) or not (−). The asterisk indicates a nonspecific cross-reacting band. Actin is shown as protein loading control.
uvr8 mutants did not respond to UV-B, the cry1 cry2 double mutant showed an enhanced activation of HY5  2 cry1 and phyA are the main photoreceptors mediating the RUP1-and RUP2-dependent blue-light repression of UVR8 activity. a Schematic representation of the light treatments of seedlings for the UVR8 re-dimerization analyses. Four-d-old etiolated seedlings were exposed to 12-h blue light (B, 50 μmol m −2 s −1 ) or maintained in darkness (D) before they were irradiated for 15 min with UV-B (0.85 mW cm −2 ) and maintained further in darkness for 1 h. The time points of sampling are indicated by red arrows labelled with −UV-B, +UV-B, 30′, and 60′. b Immunoblot analysis of UVR8 re-dimerization in wild type (Col) and rup1-1 rup2-1 (rup1 rup2) seedlings. UVR8 dimers were detectable in nonheat-denatured protein samples. Parallel denatured samples demonstrate equal amounts of UVR8 protein. The asterisk indicates a nonspecific cross-reacting band. c Quantification of the UVR8 dimer /UVR8 total ratio (%) in Col and rup1 rup2. d, e Quantification of the UVR8 dimer /UVR8 total ratio (%) in response to UV-B in Col, cry1-304 cry2-1 (cry1 cry2), phyA-211 (phyA), and cry1-304 cry2-1 phyA-412 (cry1 cry2 phyA) seedlings pre-treated with blue light (e) or not (d). f, g Quantification of the UVR8 dimer /UVR8 total ratio (%) in response to UV-B in Col, cry1-304 (cry1), cry2-1 (cry2), and cry1 cry2 seedlings pre-treated with blue light (g) or not (f). Data are means ± SEM (N = 3). *P < 0.05, **P < 0.01, and ***P < 0.001.   cry1 and phyB establish UV-B tolerance alongside UVR8. Cryptochrome blue-light and phytochrome red-light signalling are known to induce expression of genes that largely overlap with UVR8-induced genes. These co-induced genes include those associated with UV-B tolerance, such as CHS and FLS; those encoding phenylpropanoid biosynthetic enzymes, which are important for the synthesis of flavonol glycosides that function as sunscreen metabolites; and PHR1 and UVR3, which encode photolyases important for DNA damage repair 5,17,30 . This indicates that, in addition to cry1 negative regulation of UVR8 activity, cry1 as well as phyB may contribute to UV-B tolerance alongside UVR8. In the absence of UV-B, the viability of cry1, phyB, and uvr8 single mutants and cry1 uvr8, phyB uvr8, and phyB cry1 double mutants was not affected (Fig. 5). The main observable growth difference without UV-B was the long petioles of phyB plants that display a constitutive shade-response phenotype (Fig. 5), as described before 31,32 . In the presence of UV-B, cry1 and wild-type plants showed comparable growth reductions (Fig. 5). In addition to the smaller rosette phenotype under UV-B, uvr8 plants displayed chlorotic leaves but remained viable (Fig. 5).
Most strikingly, however, we observed that cry1 uvr8 plants were extremely UV-B hypersensitive and did not survive under UV-B irradiation (Fig. 5). The constitutive shade-response phenotype of phyB evident without UV-B was clearly suppressed in the presence of UV-B (Fig. 5), in agreement with the finding that UVR8 antagonizes shade-avoidance responses 33 . phyB uvr8 plant survival was also clearly impaired under UV-B, although phyB uvr8 plants were less UV-B sensitive than cry1 uvr8 (Fig. 5). Interestingly, in stark contrast to cry1 uvr8 and phyB uvr8 plants, cry1 phyB double-mutant plants maintained UV-B tolerance (Fig. 5).
We conclude that, in addition to the major role UVR8 plays in establishing UV-B acclimation and tolerance, blue light-activated cry1 and red light-activated phyB contribute to full UV-B tolerance.

Discussion
Cellular signal transduction is usually composed of positive regulatory factors that function as "accelerators" as well as their specific negative feedback regulators that function as "brakes". Indeed, plant photoreceptors are activated upon photon absorption, which in conjunction with negative feedback regulation establishes a photoequilibrium according to the prevailing light conditions. For example, cryptochromes form an equilibrium between active homodimeric and inactive monomeric states, whereas UVR8 equilibrates between an active monomeric and an inactive homodimeric state 6,17,34,35 . Thus, both types of receptors are regulated at the level of their oligomeric state, with UVR8 "photomonomerizing" and cryptochromes "photodimerizing". Moreover, both cryptochromes and UVR8 activate expression of their own negative feedback regulators, namely BIC1/BIC2 and RUP1/RUP2, respectively 13,17 . These regulatory proteins interact directly with the respective photoreceptors and facilitate their reversion to the ground state and/or inhibit their activation 13,14,17,18,34,36 (Fig. 6). Interestingly, BIC1 and BIC2 are strongly induced by UVR8 signalling, although BIC1 and BIC2 apparently do not directly function in UV-B responses. Reciprocally, RUP1 and RUP2 are strongly induced by cryptochrome signalling, although RUP1 and RUP2 do not function in responses to monochromatic blue light 13 . We show here that the UVR8 photoequilibrium is indeed responsive to cryptochromemediated blue-light signalling, thus identifying a novel photoreceptor co-action mechanism to balance UV-B sensitivity of plants under the polychromatic spectrum of sunlight (Fig. 6).
Phytochromes also mediate BIC1/BIC2 and RUP1/RUP2 transcriptional activation in response to red and far-red light 13,18 , suggesting that phytochromes contribute to the cross-regulation of cryptochromes and UVR8 through activation of their negative feedback regulators. However, the effect of phytochrome-mediated red light signalling on RUP1/RUP2 induction and UVR8 redimerization is very moderate if compared to the effect of blue light ( Supplementary Fig. 5). In phyB signalling, positive feedback regulation is mediated by the phyB-interacting proteins PHOTO-PERIODIC CONTROL OF HYPOCOTYL 1 (PCH1) and PCH1 LIKE (PCHL), which suppress phyB dark reversion from the active Pfr to the inactive Pr state 37,38 . Interestingly, in addition to red light, PCH1 and PCHL are transcriptionally activated by blue light and UV-B 5,37 . Indeed, blue light was shown to increase the response to red light facilitated by phyB 37 . Notwithstanding the physiological effects, the cross-regulation of these photoreceptor regulators is likely associated with COP1 inactivation and activation of HY5, which is the common signalling node of cryptochromes, phytochromes, and UVR8 5,9,10,[20][21][22][23]34,39,40 . The convergence of these three wavelength-specific signalling pathways on COP1-HY5 is likely the basis for their overlapping photomorphogenic responses, which include hypocotyl growth inhibition, phenylpropanoid accumulation, flowering, and gene expression changes [1][2][3]20,34,41 . In particular, as demonstrated here, phytochrome, cryptochrome, and UVR8 signalling all contribute to establishing UV-B tolerance, suggesting that plants evolved photoreceptor cross-regulation mechanisms to globally balance their light responses in nature.
In contrast to the single mutants, cry1 uvr8 and phyB uvr8 double mutants were not viable under UV-B in our growth chamber experiments, demonstrating functional redundancies between cry1 and UVR8 as well as between phyB and UVR8. Interestingly, however, the cry1 phyB double mutant could tolerate UV-B exposure, suggesting a major role for UVR8 even in the absence of both phyB and cry1. The potential contributions of the additional cryptochrome (cry2) and four phytochromes (phyA and phyC-E) remain to be determined. Moreover, phototropin 2 (phot2) plays a major role in chloroplast avoidance in response to high light 42,43 . As UV-B levels are high on sunny days and UV-B affects the photosynthetic machinery in chloroplasts 44 probable roles for additional photoreceptors in UV-B tolerance, recent work showed that triple-mutant plants lacking cry1, cry2, and UVR8 did not survive under natural sunlight, whereas they survived well when UV-B was specifically filtered out 25 . Thus, cryptochromes and UVR8 are indeed required for survival under natural conditions. It is of note that our data provide evidence that the UVR8 signalling pathway is hyperactive in cry1 cry2 mutants and BIC1/BIC2 overexpression lines. In contrast, the absence of enhanced RUP1 and RUP2 induction and associated lack of a clear effect on UVR8 activity in bic1 bic2 indicates that endogenous BIC1 and BIC2 do not play a major role in inhibiting cry activation under the conditions employed, including developmental stage, fluence rate, and kinetics. Notwithstanding this, the hyperactivity of UVR8 signalling in cry1 cry2 suggests that enhanced UVR8 signalling contributes to UV-B tolerance in the cryptochrome mutants. As red light also induces RUP1 and RUP2 expression, although to an apparently much lower level, a similar argument can be made for enhanced UVR8 signalling in phyB mutants. Moreover, as UVR8 signalling induces BIC1 and BIC2 expression, enhanced cryptochrome signalling in uvr8 single mutants may contribute to their UV-B tolerance; however, such a response remains to be shown. Thus, interactions within the network of phytochrome, cryptochrome, and UVR8 signalling pathways contribute to UV-B tolerance, allowing plant survival in sunlight in the natural environment. These cross-pathway interactions involve the commonly targeted COP1 E3 ubiquitin ligase and the upregulation of other photoreceptor-specific inhibitors. An exact understanding of these interactions and the individual contributions of the 13 Arabidopsis photoreceptors will provide a full appreciation of the roles photoreceptors play in plant survival. It is evident that RUP1 and RUP2 present a node for the integration of different light qualities allowing cross-regulation of the UVR8 activity state, which thus adapts UV-B tolerance strategies to the ambient light environment. It can also be speculated that other signalling pathways triggered by environmental signals beyond light perception may affect RUP1 and RUP2, thereby providing further cross-regulation of UVR8-mediated photomorphogenesis and UV-B acclimation.
Arabidopsis seeds were surface sterilized with sodium hypochlorite and sown on half-strength Murashige and Skoog basal salt medium (MS; Duchefa) containing 1% (w/v) agar (Applichem) supplemented with 1% (w/v) of sucrose, except for in those experiments involving etiolated seedlings. After 2-d stratification at 4°C, seedlings were grown in white light or seeds were exposed to 6-h white light to induce germination before being grown in darkness at 22°C. Quantitative real-time PCR. Plant total RNA was isolated with the Plant RNeasy kit including DNase treatment according to the manufacturer's instructions (Qiagen). Synthesis of cDNA was performed using the TaqMan Reverse Transcription Reagents kit according to the manufacturer's standard protocol (Thermo Fisher Scientific). Each quantitative real-time PCR (qRT-PCR) reaction contained cDNA synthesized with a 1:1 mixture of oligo(dT) primers and random hexamers from 300 ng of total RNA. qRT-PCR was performed using a PowerUp SYBR Green  Fig. 6 Working model for the molecular integration of UVR8 and cryptochrome signalling. Light activation of UVR8 and cry1 leads to inhibition of the COP1-SPA E3 ubiquitin ligase complex, which releases HY5 from its repression through ubiquitination and proteasomal degradation. The transcriptional activator HY5 induces the expression of genes such as those associated with UV-B acclimation and tolerance (biosynthesis of sunscreen metabolites, DNA repair, photoprotection, etc.). HY5 also induces genes encoding distinct proteins with regulatory roles that concurrently provide feedback regulation of UVR8 and cryptochrome photoreceptor activities (note that the control of cry1 activity by UV-B through UVR8 signalling remains to be demonstrated). BIC1 and BIC2 are thought to mainly block cry1 homodimerization, whereas RUP1 and RUP2 are thought to primarily facilitate UVR8 re-dimerization (although additionally blocking UVR8 monomerization cannot be excluded).
Anthocyanin extraction and measurement. Plant seedlings were grown 4 d in the same conditions as for hypocotyl length measurement, and 50 mg plant material was homogenized for 10 s using a Silamat S5 mixer (Ivoclar Vivadent). 250 μl of acidic methanol (1% HCl) was added to each sample and placed in an overhead shaker at 4°C for 1 h. Samples were centrifuged for 1 min at 14,000 rpm and the supernatant was used to quantify anthocyanin content based on the absorption at 535 and 650 nm. Values were reported as A 535 − 0.25 (A 650 ) g −1 fresh weight.
Statistical analyses. Statistical analyses were performed using GraphPad Software Prism 8 software (San Diego, California). Statistical significance of the differences between means was determined using Student's t tests (protein quantification data) or one-way ANOVA and post-hoc Tukey tests (hypocotyl length data).

Data availability
The data supporting the findings in this study are available and described within the paper and its supplementary files. The source data underlying Figs. 1-5 and Supplementary Figs. 1−5 are provided as a separate Source Data file. All biological material generated in this study are available from the corresponding author upon reasonable request.