Reversion of the Arabidopsis UV-B photoreceptor UVR8to the homodimeric ground stateMarc Heijde and Roman Ulm1

Department of Botany and Plant Biology, Sciences III, University of Geneva, CH-1211 Geneva 4, Switzerland

Edited by George Coupland, Max Planck Institute for Plant Breeding Research, Cologne, Germany, and approved December 12, 2012 (received for reviewAugust 16, 2012)

Plants require the UV-B photoreceptor UV RESISTANCE LOCUS 8(UVR8) for acclimation and survival in sunlight. Upon UV-B percep-tion, UVR8 switches instantaneously from a homodimeric tomonomeric configuration, which leads to interaction with the keysignaling protein CONSTITUTIVELY PHOTOMORPHOGENIC 1 (COP1)and induction of UV-B–protective responses. Here, we show thatUVR8 monomerization is reversible in vivo, restoring the homo-dimeric ground state. We also demonstrate that the UVR8-inter-acting proteins REPRESSOR OF UV-B PHOTOMORPHOGENESIS(RUP)1 and RUP2 mediate UVR8 redimerization independently ofCOP1. UVR8 redimerization consequently disrupts the UVR8–COP1interaction, which halts signaling. Our results identify a key roleof RUP1- and RUP2-mediated UVR8 redimerization in photore-ceptor inactivation, a crucial process that regenerates reactivat-able UVR8 homodimers.

light signaling | photobiology | signal transduction

UV-B radiation (UV-B; 280–315 nm) is an integral part ofsunlight with a strong impact on terrestrial ecosystems (1–3).

In plants, UV-B perception is necessary for UV-B acclimationand UV-B stress tolerance (4–6). Specific UV-B perception isfacilitated by the photoreceptor UV RESISTANCE LOCUS 8(UVR8) identified only recently in Arabidopsis (7). In agreementwith its photoreceptor function, uvr8 null mutants show a stronglyreduced response to UV-B (8–11), which even is absent underconditions specifically activating UV-B photoreceptor responses(4). In contrast, UV-B stress responses are not affected per se inuvr8 mutants (12).Upon UV-B irradiation, UVR8 homodimers monomerize in-

stantaneously to active monomers (7). The UVR8 monomer theninteracts with the WD40-repeat domain of the E3 ubiquitin ligaseCONSTITUTIVELY PHOTOMORPHOGENIC 1 (COP1) (4),a central regulator of light-dependent plant photomorphogenesisand also of utmost importance in UV-B signaling (13, 14). COP1–UVR8 interaction is an early event in the UV-B perception andsignaling pathway and essential for UV-B–dependent photo-morphogenesis and acclimation (4). One of the main molecularoutcomes of this interaction is an increase in protein level of thebZIP transcription factor ELONGATED HYPOCOTYL 5(HY5), which may be the result of reduced HY5 ubiquitinationby COP1 (4). HY5 together with its homolog HYH induceexpression of the majority but not all genes included in theUVR8-dependent UV-B response (15–18). In a negative feed-back loop, the light-regulated SALT TOLERANCE/B-BOXDOMAIN PROTEIN 24 (STO/BBX24) was shown to fine-tunethe UV-B response by impinging on HY5 (19).UVR8 is a seven-bladed β-propeller protein that makes use of

tryptophan residues intrinsic to the protein as chromophores forUV-B absorption, with a primary role established for tryptophan-285 (7, 20, 21). In agreement with the major role that Trp-285plays in UV-B–mediated monomerization of UVR8 (7), it wasproposed that UV-B absorption by specific tryptophans, namelyTrp-285 and Trp-233, leads to disruption of cross-dimer salt bridgesinvolving crucial arginins (20, 21). Despite recent progress indescribing UVR8 monomerization and activation of UV-B

signaling, mechanisms behind in vivo UVR8 inactivation re-main poorly understood.We recently described theWD40-repeat proteins REPRESSOR

OF UV-B PHOTOMORPHOGENESIS (RUP)1 and RUP2as negative feedback regulators of the UV-B–signaling cascade(22). Upon UV-B exposure, the RUP1 and RUP2 genes aretranscriptionally activated in a UVR8-dependent manner. RUP1-and RUP2-YFP fusion proteins localize to both the nucleus andthe cytoplasm (22), mimicking the subcellular localization of UVR8(23). RUP1 and RUP2 are known to repress the UV-B–signalingpathway, but the mechanism by which they do so is presentlyunknown (22). However, direct interaction of RUP1 and RUP2with UVR8 suggests that their repressive mechanism is at thephotoreceptor level (22).In the present study, we demonstrate that the UVR8 photo-

receptor is capable of in vivo redimerization, restoring thehomodimeric ground state, and that this process requires RUP1and RUP2, but is not affected by the presence or absence ofCOP1. We further provide evidence that RUP1- and RUP2-mediated UVR8 redimerization results in the disruption ofUVR8–COP1 interaction. The UVR8 “off switch” mechanismthus uses specific regulatory proteins to mediate reversion ofUVR8 from the signaling to the ground state by redimerization,a process that is of major importance for optimal plant growthand development in sunlight.

Results and DiscussionUV-B–Dependent UVR8 Monomerization Is Reversible in Vivo. Tounderstand UVR8 protein dynamics following UV-B perception,we investigated reversion of the UVR8 monomer back to its di-mer conformation. Inactive UVR8 homodimers can be detectedon protein gel blots of non–heat-denatured protein samples (7).Following UV-B–dependent monomerization, UVR8 redimeri-zation was apparent already 30 min post UV-B exposure, andcomplete redimerization was observed within approximately 2 h(Fig. 1A). Protein gel blots created in parallel using heat-dena-tured aliquots of the same protein samples showed comparablelevels of UVR8, demonstrating that UVR8 levels remain stableduring the recovery phase (Fig. 1A). This in vivo UVR8 redime-rization is much faster than recent reports of more than 24 hrequired for the completion of redimerization in vitro (20, 21).To rule out the possibility that reappearance of the UVR8homodimer is dependent on de novo synthesis of UVR8, weanalyzed UVR8 protein dynamics in the presence of cyclohexi-mide (CHX). Under conditions where CHX efficiently blockedprotein translation (Fig. S1), clear UVR8 redimerization wasobserved at a rate comparable to the mock treatment (Fig. 1A,

Author contributions: M.H. and R.U. designed research; M.H. performed research; M.H.and R.U. analyzed data; and M.H. and R.U. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.1To whom correspondence should be addressed. E-mail: roman.ulm@unige.ch.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1214237110/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1214237110 PNAS | January 15, 2013 | vol. 110 | no. 3 | 1113–1118

PLANTBIOLO

GY

Dow

nloa

ded

by g

uest

on

Nov

embe

r 8,

202

0

Right). To investigate the monomerization or “reactivation” po-tential of redimerized UVR8, we subjected plants to up to threesuccessive UVR8-monomerizing UV-B treatments with a redi-merization recovery period of 4 h between each treatment.UVR8 dimers repeatedly monomerized in response to UV-B andredimerized in its absence, regardless of preceding cycles ofmonomerization/redimerization and also in the presence ofCHX (Fig. 1B). We thus conclude that UVR8 redimerization isa property of the protein that is independent of UVR8 de novosynthesis, allowing the regeneration of photoactive homodimers.In addition, UVR8 protein dynamics are fully reversible in vivowith monomerization and dimerization depending on the pres-ence and absence of UV-B, respectively.

RUP1 and RUP2 Regulate the UVR8 Dimer-to-Monomer Ratio. RUP1and RUP2 are negative feedback regulators of UVR8, whichfunction upstream or adjacent to the UVR8–COP1 interaction(22). The RUP1 and RUP2 mechanism of action may be bypreventing UVR8 monomerization and/or facilitating UVR8redimerization post UV-B exposure. Consistent with both pos-sibilities and in contrast to the UV-B–specific UVR8–COP1 in-teraction, RUP1 and RUP2 can interact with UVR8 underconditions with and without UV-B and thus with both UVR8monomers and homodimers (22, 24). To begin with, we testedwhether RUP1 and RUP2 influence the UVR8 dimer–monomerratio upon UV-B exposure, which would imply that RUP1 andRUP2 act upstream of the UVR8–COP1 interaction. Under low-level UV-B where UVR8 monomerization is almost undetect-able in wild type, clear UVR8 monomerization was apparent inthe rup1 rup2 double mutant (Fig. 2A). This is in agreement withthe UV-B hyper-responsiveness and enhanced acclimation toUV-B in rup1 rup2 (22). Conversely, under UV-B irradiation thatefficiently monomerizes UVR8 in wild type, UVR8 monomer-ization in a RUP2-overexpression line was undetectable (Fig. 2B).Again, this result is in agreement with the UV-B hypo-respon-siveness and reduced acclimation to UV-B that accompaniesRUP2 overexpression (22). Moreover, it is of note that the cop1-4

mutant—which expresses the truncated COP1N282 protein thatlacks the WD40-repeat domain and thus cannot interact withUVR8 (4, 7)—did not show any difference in comparison withthe wild type (Fig. 2 A and B). This suggests that COP1, andtherefore also the UVR8–COP1 interaction, has no major role inthe regulation of UVR8 activity. Total levels of UVR8 in therup1 rup2 double mutant and the RUP2-overexpression line werecomparable to wild type, suggesting that the effect on UVR8monomerization was not due to variations in UVR8 levels (Fig. 2A and B; heat-denatured samples).As mentioned above, RUP1 and RUP2 may negatively regu-

late UV-B signaling by mediating UVR8 redimerization postUV-B exposure. This is supported by the fact that RUP1 andRUP2 levels are induced by UV-B exposure to then providenegative feedback regulation of the UVR8 pathway involvingdirect RUP1/RUP2–UVR8 interaction (22). Thus, we furtherinvestigated whether the level of RUP1 and RUP2 influencesUVR8 redimerization post UV-B exposure. First, we analyzedUVR8 protein dynamics in the absence of RUP1 and RUP2.Although wild-type seedlings showed complete UVR8 dimerrecovery within 2 h post UV-B exposure (Fig. 1A), we observedthat rup1 rup2 double mutants were strongly impaired in UVR8dimer recovery during the entire 6-h duration of the experiment(Fig. 2C). Even 36 h post UV-B exposure UVR8 redimerizationwas not yet completed in rup1 rup2, even though redimerizationduring that time became apparent in the absence of RUP1 andRUP2 (Fig. S2). Together, this clearly shows that, without RUP1and RUP2, the ability of monomerized UVR8 to revert to theinactive homodimeric state is compromised. Relatively normalUVR8 redimerization was seen in rup1 and rup2 single mutants(Fig. S3), reemphasizing the functional redundancy of RUP1 andRUP2 in regulating UVR8 (22). We further tested whetherRUP1 and RUP2 may also partially inhibit UVR8 monomerformation. However, UVR8 monomerization within minutes ofUV-B was similar in wild-type and rup1 rup2 double mutants(Fig. S4). Although we cannot formally exclude that the presenceof RUP1 and RUP2 also prevents UV-B–dependent UVR8

Mock

-UV +UV 30’ 1h 2h 4h 6h

+ CHX

-UV +UV 30’ 1h 2h 4h 6h

- actin

A

B

-UV +UV 4h +UV 4h +UV

Mock + CHX

-UV +UV 4h +UV 4h +UV

Non-denatured

Denatured

- UVR8

- UVR8

- UVR8dimer

Non-denatured

Denatured

- UVR8

- UVR8

- UVR8dimer

*

Denatured

Recovery in WLRecovery in WL

Fig. 1. UV-B–activated UVR8 monomers redimerize in white light without UV-B, and regenerated UVR8 dimers are competent for UV-B perception andmonomerization. (A) Seven-day-old seedlings were irradiated by broadband UV-B for 15 min (+UV) to allow UVR8 monomerization before they weresubjected to a recovery period in white light (WL) for the indicated times. An asterisk (*) indicates a nonspecific cross-reacting band. (B) Seven-day-oldseedlings were repeatedly treated for 15 min with broadband UV-B (+UV) and allowed to recover for 4 h in white light (4 h) (i.e., from left to right: −UV/+UV/+UV+4h/+UV+4h+UV/+UV+4h+UV+4h/+UV+4h+UV+4h+UV). (A and B) Mock-treated plants were in liquid MS medium with DMSO, and cycloheximide(CHX)-treated plants (+CHX) were in liquid MS medium supplied with 0.5 mM CHX in DMSO. UVR8 dimers were detectable in non–heat-denatured proteinsamples, as described before (7). Parallel denatured samples demonstrated equal amounts of UVR8 protein.

1114 | www.pnas.org/cgi/doi/10.1073/pnas.1214237110 Heijde and Ulm

Dow

nloa

ded

by g

uest

on

Nov

embe

r 8,

202

0

monomerization to a minor extent, it is clear that RUP1 andRUP2 play a major role in facilitating UVR8 redimerization.Second, we analyzed UVR8 protein dynamics with an increasedlevel of RUP2. In contrast to the clear transition of UVR8 tomonomer in wild-type and rup1 rup2 double mutants upon UV-Bexposure, no change in the UVR8 homodimer/monomer ratiowas observed in a RUP2-overexpression line (Fig. 2 B and D).This suggests that high RUP2 levels lead to instantaneous re-covery of UVR8 homodimer.UVR8 protein dynamics in rup1 rup2 and RUP2-overexpression

lines is in agreement with the physiological role of RUP1 andRUP2 as negative feedback regulators through direct interactionwith UVR8 (22). Therefore, although UV-B–dependent UVR8monomerization is an intrinsic property of the protein itself (7, 20,21), we conclude that the efficient reversion of UVR8 monomerto the inactive homodimeric ground state in vivo is strongly de-pendent on RUP1 and RUP2. Thus, we propose that UVR8redimerization is the major mechanism of action of RUP1 andRUP2 as negative feedback regulators of UV-B signaling.

RUP1 and RUP2 Negatively Regulate UVR8–COP1 Interaction. Fol-lowing UV-B perception by UVR8, a primary molecular event keyto the UV-B–signaling pathway is direct interaction of the UVR8monomer with COP1 (4, 7). It was thus of interest to see whetherthe UVR8–COP1 interaction was affected by RUP1 and RUP2.We have generated anti-COP1 and anti-UVR8 antibodies thatallow UV-B–dependent coimmunoprecipitation (co-IP) of en-dogenous COP1 with UVR8 from wild-type seedlings (Fig. 3).Interestingly, under a level of UV-B that is insufficient to producea detectableUVR8–COP1 interaction in wild type, COP1 could beco-IP’ed with UVR8 in rup1 rup2 double mutants (Fig. 3A). Also,

under a level of UV-B that yields clear co-IP of COP1 with UVR8in wild type, elevated levels of COP1 could be co-IP’ed withUVR8in rup1 rup2 double mutants (Fig. 3B). In contrast, no COP1 co-IPcould be detected in a RUP2-overexpression line (Fig. 3B). Ourdata demonstrate that the UVR8–COP1 interaction is stronglyinfluenced by RUP1 and RUP2. Thus, negative regulation of theUVR8–signaling pathway by RUP1 and RUP2 is associated withthe disruption of UVR8–COP1 interaction.We further investigated the stability of the UVR8–COP1 in-

teraction post UV-B exposure to better understand how thisfacet of UVR8 signaling was inactivated. Following UV-B irra-diation that leads to UVR8–COP1 interaction, we monitored theUVR8–COP1 interaction over a 4-h recovery period whereseedlings were incubated under white light devoid of UV-B. Inagreement with the timing of UVR8 redimerization (Fig. 1A),reduced UVR8–COP1 interaction was apparent already 30 minpost UV-B exposure in wild-type seedlings (Fig. 3C). No furtherCOP1 co-IP was detectable after 2 h white light incubation (Fig.3C). In stark contrast, the UVR8–COP1 interaction was barelyaffected in rup1 rup2 double mutants after transfer from sup-plemental UV-B to white light, with clear COP1 co-IP main-tained for up to 4 h post UV-B exposure (Fig. 3D). We concludethat UVR8 redimerization, heavily influenced by the action ofRUP1 and RUP2, disrupts the UVR8–COP1 interaction. Thishighlights the underlying mechanism of RUP1 and RUP2 asrepressors of the UVR8–signaling pathway.

RUP1 and RUP2 Regulate UVR8 Redimerization Independently ofCOP1. Thus far, evidence has been provided for the role of RUP1and RUP2 in UVR8 redimerization leading to disruption of theUVR8–COP1 interaction. However, the underlying interaction

A

rup1 rup2

-UV +UV 30’ 1h 2h 4h 6h

RUP2 Ox#3

-UV +UV 30’ 1h 2h 4h 6h

B

DC

Non-denatured

- + - + - + - +

Denatured

- UVR8

- UVR8

- UVR8dimer

UV-B- UV-B- + - + - + - +- UVR8dimer

- UVR8

- UVR8Non-denatured

Denatured

-

Non-denatured

Denatured

- UVR8

- UVR8

- UVR8dimer

Non-denatured

Denatured

- UVR8

- UVR8

- UVR8dimer

**

Recovery in WL Recovery in WL

Fig. 2. Redimerization of UVR8 is mediated by RUP1 and RUP2. (A) Seven-day-old seedlings were irradiated for 6 h with supplementary narrowband UV-B (+)or without (−). Wild type (Col) was compared with rup1 rup2, rup2-1/Pro35S:RUP2 (RUP2 Ox#3), cop1-4, and uvr8-6. (B) Seven-day-old seedlings were irradiatedfor 6 h with supplementary narrowband UV-B and 15-min broadband UV-B (+) or without (−). (C and D) Seven-day-old rup1 rup2 (C) or rup2-1/Pro35S:RUP2(RUP2 Ox#3) (D) seedlings were irradiated for 15 min with broadband UV-B before recovery in white light (WL) for the indicated times. An asterisk (*)indicates a nonspecific cross-reacting band. (A–D) UVR8 dimers were detectable in non–heat-denatured protein samples, as described before (7). Paralleldenatured samples demonstrated equal amounts of UVR8 protein.

Heijde and Ulm PNAS | January 15, 2013 | vol. 110 | no. 3 | 1115

PLANTBIOLO

GY

Dow

nloa

ded

by g

uest

on

Nov

embe

r 8,

202

0

mechanism between these four proteins and the primary effect ofRUP1 and RUP2 remains unknown. We initially hypothesizedthat interaction with COP1 stabilized the UVR8 monomer andthat UVR8 redimerization is due to direct disruption of theUVR8–COP1 interaction by RUP1 and RUP2. To test this, weanalyzed UVR8 redimerization in the absence of functionalCOP1. We generated cop1 rup1 rup2 triple mutants and com-pared UVR8 protein dynamics to that of cop1 single mutants,expecting to see comparable rates of UVR8 redimerization if ourhypothesis held true. However, UVR8 redimerization was muchslower in the cop1 rup1 rup2 triple mutant than in the cop1mutant, clearly demonstrating that RUP1 and RUP2 mediateUVR8 redimerization even in the absence of COP1 (Fig. 4).Indeed, UVR8 redimerization was comparable in both cop1 andwild type, whereas it was notably impaired in rup1 rup2 and cop1rup1 rup2 (Fig. 4). Combined, these data argue against a modelin which RUP1- and RUP2-mediated disruption of the UVR8–COP1 interaction allows UVR8 redimerization. We thus proposean alternate model in which RUP1 and RUP2 mediate UVR8redimerization independently of COP1, and it is this process thatreleases COP1. It remains to be determined whether the releaseof COP1 is solely due to UVR8 redimerization or whethercompetition between RUP1/RUP2 and COP1 for binding siteswith UVR8 may play a role as well. The fact that inactive UVR8dimers do not interact with COP1 (7) supports the hypothesisthat the redimerization may be sufficient to impact on theUVR8–COP1 interaction.

Conclusions. Light-induced changes in the structural conformationof plant photoreceptors is a common mechanism to “switch on”

corresponding light-signaling pathways. Likewise, reconversion ofthe photoreceptor to its original conformation provides the signal-ing “off switch.” In the absence of UV-B, UVR8 exists as a homo-dimer held together by a large number of intermolecular hydrogenbonds that are disrupted followingUV-B absorption by Trp-285 andTrp-233 leading to UVR8 monomerization (7, 20, 21). Here wedemonstrate that RUP1 and RUP2 play a major role in the reversalof this process. The mechanistic and structural basis of how RUP1and RUP2 mediate UVR8 redimerization to form inactive yetfunctional UVR8 homodimers awaits further investigation. Not-withstanding this, our present data clearly demonstrate that RUP1and RUP2 provide negative regulation of the UVR8-mediated UV-B–signaling pathway by mediating UVR8 redimerization, which isof primary importance to optimally balance the UV-B–inducedphotomorphogenic response with plant growth and development.

Materials and MethodsPlant Material and Growth Conditions. The cop1-4, uvr8-6, and rup1-1 rup2-1mutants are in the Columbia (Col) accession (4, 22, 25). The RUP2-over-expression line RUP2Ox#3 is in the rup2-1 background (rup2-1/Pro35S:RUP2)(22). Arabidopsis seeds were surface-sterilized and sown on half-strengthMurashige and Skoog basal salt medium (MS; Duchefa) containing 1%(wt/vol) sucrose and 1% (wt/vol) phytagel (Sigma). Seeds were stratifiedfor 2 d at 4 °C and germinated at 22 °C in a standard growth chamber underconstant white light.

For CHX experiments, seedlings were transferred to liquid half-strengthMSmedium containing 1% (wt/vol) sucrose for 15 h before 0.5 mM CHX in DMSO(Sigma) was added. For mock controls, an equivalent volume of DMSOwas added.

UV-B treatments were performed using previously established conditionswith broadband (Philips TL40W/12RS; 21 μmol·m−2·s−1) (17) or narrowband

BA

DC

-UV +UV 5’ 15’ 30’ 1h 2h 4h

Recovery in WL

rup1 rup2

- UVR8

- COP1

- UVR8

- actin

Inpu

tIP

: -U

VR

8-UV +UV 5’ 15’ 30’ 1h 2h 4h

Recovery in WL

Col

Inpu

tIP

: -U

VR

8

- UVR8

- COP1

- UVR8

- actin

- actin

- + - + - + + +- UVR8

- COP1

- UVR8

Inpu

tIP

: -U

VR

8 UV-B

- actin

- UVR8

- COP1

- UVR8

Inpu

tIP

: -U

VR

8

- + - + - + + + UV-B

Fig. 3. UVR8 interaction with COP1 is negatively regulated by RUP1 and RUP2. (A and B) Coimmunoprecipitation of COP1 using UVR8 antibodies in extracts from7-d-old wild-type (Col), rup1 rup2, RUP2 Ox#3, uvr8-6, and cop1-4 seedlings. (A) Seedlings were irradiated for 6 h with narrowband UV-B (+) or without (−).(B) Seedlings were irradiated for 6 h with supplemental narrowband UV-B and 15-min broadband UV-B (+) or without (−). (C and D) Coimmunoprecipi-tation of COP1 using UVR8 antibodies in extracts from wild-type (Col) (C ) and rup1 rup2 double mutant (D). Seven-day-old seedlings were treated with 6 hnarrowband UV-B and 15-min broadband UV-B, followed by recovery in white light (WL) for the indicated time.

1116 | www.pnas.org/cgi/doi/10.1073/pnas.1214237110 Heijde and Ulm

Dow

nloa

ded

by g

uest

on

Nov

embe

r 8,

202

0

UV-B lamps (Philips TL20W/01RS; 1.5 μmol·m−2·s−1) (13), as indicated. Al-though both treatments activate UVR8 photoreceptor-dependent UV-B re-sponses, broadband irradiation allows very efficient UVR8 monomerizationwith short-term irradiation (7), whereas narrowband irradiation allows long-term irradiation for physiological responses, but includes the activation ofnegative feedback regulation in parallel (4, 22).

Antibodies. Rabbit polyclonal antibodies were generated against syntheticpeptides derived from the UVR8 protein sequence [amino acids 410–424 + C:VPDETGLTDGSSKGNC; anti-UVR8(410–424)] and the COP1 protein sequence[amino acids C + 13–26: CVKPDPRTSSVGEGA; anti-COP1(13–26)] and were af-finity-purified against the peptide (Eurogentec). Guinea pig polyclonalantibodies were generated against a synthetic peptide [amino acids C + 426–440: CGDISVPQTDVKRVRI; anti-UVR8(426–440)] already previously used to gen-erate polyclonal UVR8 antibodies in rabbits (4) and were affinity-purifiedagainst the peptide (Eurogentec).

Protein Immunoprecipitation. Proteins were extracted in extraction buffer [50mM Tris, pH 7.5, 150 mMNaCl, 10% (vol/vol) glycerol, 5 mMMgCl2, 0.1% (vol/vol) Igepal, 2 mM benzamidine, 10 μM dichloroisocumarin, 1% (vol/vol)protease inhibitor mixture for plant extracts (Sigma), 10 μM MG132]. Im-munoprecipitation of UVR8 was performed using anti-UVR8(426–440) poly-clonal rabbit antibodies (4) for 2 h, and immunoprecipitate was capturedwith protein A-agarose (Roche Applied Science) for 1 h.

Protein Gel Blot Analysis. For protein gel blot analysis, total cellular proteins(15 μg) or immunoprecipitates were separated by electrophoresis in 8% (wt/vol)SDS–polyacrylamide gels and electrophoretically transferred to PVDFmembranesaccording to the manufacturer’s instructions (Bio-Rad). We used anti-COP1(13–26),anti-UVR8(426–440), anti-actin (A0480; Sigma-Aldrich), and anti-chalcone syn-thase (sc-12620; Santa Cruz Biotechnology) as primary antibodies, with HRP-conjugated protein A (Pierce) or anti-rabbit, anti-goat, and anti-mouseimmunoglobulins (DAKO) used as secondary antibodies, as required. Weused anti-UVR8(426–440) from guinea pig and HRP-conjugated anti-guinea pigsecondary antibodies (ab7139; Abcam) to detect UVR8 in immunoprecipi-tates generated using anti-UVR8(426–440) from rabbits. Signal detection wasperformed using the ECL Plus Western Detection Kit (GE Healthcare).

Analysis of UVR8 dimers was conducted as follows: proteins were ex-tracted in 50mM Tris, pH 7.6, 150mMNaCl, 2 mM EDTA, 1% (vol/vol) Igepal(Sigma), 1% (vol/vol) protease inhibitor mixture for plant extracts (Sigma),10 μM MG132, and 10 μM N-Acetyl-L-leucyl-L-leucyl-L-norleucinal (ALLN).Twenty micrograms of proteins were separated by electrophoresis in 8%(wt/vol) SDS–polyacrylamide gels. Gels were UV-B–irradiated before elec-trophoretic transfer to PVDF membrane, as described previously (7). Blottingwas performed using rabbit anti-UVR8(410–424) antibodies.

ACKNOWLEDGMENTS. We are grateful to Kimberley Tilbrook and MichelGoldschmidt-Clermont for critically reading the manuscript. This study wassupported by the University of Geneva and the Swiss National ScienceFoundation (Grant 31003A_132902).

1. Ballaré CL, Caldwell MM, Flint SD, Robinson SA, Bornman JF (2011) Effects of solar

ultraviolet radiation on terrestrial ecosystems. Patterns, mechanisms, and interactions

with climate change. Photochem Photobiol Sci 10(2):226–241.2. Paul ND, Gwynn-Jones D (2003) Ecological roles of solar UV radiation: Towards an

integrated approach. Trends Ecol Evol 18(1):48–55.3. Jansen MAK, Gaba V, Greenberg BM (1998) Higher plants and UV-B radiation: Bal-

ancing damage, repair and acclimation. Trends Plant Sci 3(4):131–135.4. Favory JJ, et al. (2009) Interaction of COP1 and UVR8 regulates UV-B-induced

photomorphogenesis and stress acclimation in Arabidopsis. EMBO J 28(5):591–

601.

5. Heijde M, Ulm R (2012) UV-B photoreceptor-mediated signalling in plants. Trends

Plant Sci 17(4):230–237.6. Jenkins GI (2009) Signal transduction in responses to UV-B radiation. Annu Rev Plant

Biol 60:407–431.7. Rizzini L, et al. (2011) Perception of UV-B by the Arabidopsis UVR8 protein. Science

332(6025):103–106.8. Brown BA, et al. (2005) A UV-B-specific signaling component orchestrates plant UV

protection. Proc Natl Acad Sci USA 102(50):18225–18230.9. Kliebenstein DJ, Lim JE, Landry LG, Last RL (2002) Arabidopsis UVR8 regu-

lates ultraviolet-B signal transduction and tolerance and contains sequence

-UV +UV 30’ 1h 2h 4h 6h

ColRecovery in WL

-UV +UV 30’ 1h 2h 4h 6h

rup1 rup2

Recovery in WL

-UV +UV 30’ 1h 2h 4h 6h

rup1 rup2 cop1-4

Recovery in WL

Non-denatured

Denatured

- UVR8

- UVR8

- UVR8dimer

*

-UV +UV 30’ 1h 2h 4h 6h

cop1-4

Recovery in WL

Non-denatured

Denatured

- UVR8

- UVR8

- UVR8dimer

*

UVR8

UVR8

UVR8

RUP1

RUP2

UV-B

COP1

UVR8

UVR8

UVR8

UV-B

COP1RUP1

RUP2

UVR8

UVR8

UVR8

RUP1

RUP2

UV-B

COP1

UVR8

UVR8

UVR8

RUP1

RUP2

UV-B

COP1

Fig. 4. RUP1- and RUP2-mediated redimerization of UVR8 is independent of COP1. Seven-day-old seedlings were irradiated for 15 min with broadband UV-Bbefore recovery in white light (WL) for the indicated times. UVR8 dimers were detectable in non–heat-denatured protein samples, as described before (7).Parallel denatured samples demonstrated equal amounts of UVR8 protein. An asterisk (*) indicates a nonspecific cross-reacting band.

Heijde and Ulm PNAS | January 15, 2013 | vol. 110 | no. 3 | 1117

PLANTBIOLO

GY

Dow

nloa

ded

by g

uest

on

Nov

embe

r 8,

202

0

similarity to human regulator of chromatin condensation 1. Plant Physiol 130(1):234–243.

10. Wargent JJ, Gegas VC, Jenkins GI, Doonan JH, Paul ND (2009) UVR8 in Arabidopsisthaliana regulates multiple aspects of cellular differentiation during leaf developmentin response to ultraviolet B radiation. New Phytol 183(2):315–326.

11. Demkura PV, Ballaré CL (2012) UVR8 mediates UV-B-induced Arabidopsis defenseresponses against Botrytis cinerea by controlling sinapate accumulation. Mol Plant5(3):642–652.

12. González Besteiro MA, Bartels S, Albert A, Ulm R (2011) Arabidopsis MAP kinasephosphatase 1 and its target MAP kinases 3 and 6 antagonistically determine UV-B stresstolerance, independent of the UVR8 photoreceptor pathway. Plant J 68(4):727–737.

13. Oravecz A, et al. (2006) CONSTITUTIVELY PHOTOMORPHOGENIC1 is required for theUV-B response in Arabidopsis. Plant Cell 18(8):1975–1990.

14. Yi C, Deng XW (2005) COP1: From plant photomorphogenesis to mammalian tu-morigenesis. Trends Cell Biol 15(11):618–625.

15. Brown BA, Jenkins GI (2008) UV-B signaling pathways with different fluence-rateresponse profiles are distinguished in mature Arabidopsis leaf tissue by requirementfor UVR8, HY5, and HYH. Plant Physiol 146(2):576–588.

16. Fehér B, et al. (2011) Functional interaction of the circadian clock and UV RESISTANCELOCUS 8-controlled UV-B signaling pathways in Arabidopsis thaliana. Plant J 67(1):37–48.

17. Ulm R, et al. (2004) Genome-wide analysis of gene expression reveals function of thebZIP transcription factor HY5 in the UV-B response of Arabidopsis. Proc Natl Acad SciUSA 101(5):1397–1402.

18. Stracke R, et al. (2010) The Arabidopsis bZIP transcription factor HY5 regulates ex-

pression of the PFG1/MYB12 gene in response to light and ultraviolet-B radiation.Plant Cell Environ 33(1):88–103.

19. Jiang L, et al. (2012) Arabidopsis STO/BBX24 negatively regulates UV-B signaling by

interacting with COP1 and repressing HY5 transcriptional activity. Cell Res 22(6):1046–1057.

20. Christie JM, et al. (2012) Plant UVR8 photoreceptor senses UV-B by tryptophan-mediated disruption of cross-dimer salt bridges. Science 335(6075):1492–1496.

21. Wu D, et al. (2012) Structural basis of ultraviolet-B perception by UVR8. Nature

484(7393):214–219.22. Gruber H, et al. (2010) Negative feedback regulation of UV-B-induced photomor-

phogenesis and stress acclimation in Arabidopsis. Proc Natl Acad Sci USA 107(46):20132–20137.

23. Kaiserli E, Jenkins GI (2007) UV-B promotes rapid nuclear translocation of the Ara-

bidopsis UV-B specific signaling component UVR8 and activates its function in thenucleus. Plant Cell 19(8):2662–2673.

24. Cloix C, et al. (2012) C-terminal region of the UV-B photoreceptor UVR8 initiatessignaling through interaction with the COP1 protein. Proc Natl Acad Sci USA 109(40):

16366–16370.25. McNellis TW, et al. (1994) Genetic and molecular analysis of an allelic series of cop1

mutants suggests functional roles for the multiple protein domains. Plant Cell 6(4):

487–500.

1118 | www.pnas.org/cgi/doi/10.1073/pnas.1214237110 Heijde and Ulm

Dow

nloa

ded

by g

uest

on

Nov

embe

r 8,

202

0