HY5, a positive regulator of light signaling, negativelycontrols the unfolded protein response in ArabidopsisGanesh M. Nawkara,1, Chang Ho Kanga,1, Punyakishore Maibama,1, Joung Hun Parka, Young Jun Junga,Ho Byoung Chaea, Yong Hun Chia, In Jung Junga, Woe Yeon Kima, Dae-Jin Yuna, and Sang Yeol Leea,2

aDivision of Applied Life Sciences (BK21+) and Plant Molecular Biology and Biotechnology Research Center, Gyeongsang National University,Jinju 52828, Korea

Edited by Xinnian Dong, Duke University, Durham, NC and approved December 22, 2016 (received for review June 18, 2016)

Light influences essentially all aspects of plant growth and devel-opment. Integration of light signaling with different stress re-sponse results in improvement of plant survival rates in everchanging environmental conditions. Diverse environmental stressesaffect the protein-folding capacity of the endoplasmic reticulum(ER), thus evoking ER stress in plants. Consequently, the unfoldedprotein response (UPR), in which a set of molecular chaperones isexpressed, is initiated in the ER to alleviate this stress. Althoughits underlying molecular mechanism remains unknown, light isbelieved to be required for the ER stress response. In this study,we demonstrate that increasing light intensity elevates the ERstress sensitivity of plants. Moreover, mutation of the ELONGATEDHYPOCOTYL 5 (HY5), a key component of light signaling, leads totolerance to ER stress. This enhanced tolerance of hy5 plants canbe attributed to higher expression of UPR genes. HY5 negativelyregulates the UPR by competing with basic leucine zipper 28(bZIP28) to bind to the G-box–like element present in the ER stressresponse element (ERSE). Furthermore, we found that HY5 un-dergoes 26S proteasome-mediated degradation under ER stressconditions. Conclusively, we propose a molecular mechanism ofcrosstalk between the UPR and light signaling, mediated by HY5,which positively mediates light signaling, but negatively regulatesUPR gene expression.

endoplasmic reticulum stress | light signaling | protein-folding capacity |crosstalk | unfolded protein response

Because of their sessile lifestyle, plants have developed ad-vanced mechanisms to cope with many forms of environmental

stress conditions. Under changing environmental conditions, theadaptive responses of plants largely depend on the proper inte-gration of light signaling and stress response pathways. Plants areequipped with multiple photoreceptors, including red/far-redlight-absorbing phytochromes, blue/UV (UV)-A light-absorbingcryptochromes and phototropins, and the UV-B light-absorbingUV RESISTANCE LOCUS 8 (1, 2). The downstream componentELONGATED HYPOCOTYL 5 (HY5), a basic leucine zipper(bZIP) transcription factor (TF), mediates photoreceptor re-sponses to promote photomorphogenesis (3). Recently, genome-wide gene expression analyses and chromatin immunoprecipitation(ChIP) studies have shown that HY5 is a higher hierarchical reg-ulator of transcriptional networks for photomorphogenesis (4, 5).In particular, HY5 binds to the promoter of light-responsivegenes featuring “ACGT-containing elements” such as the G-box(CACGTG), C-box (GACGTC), Z-box (ATACGGT), and A-box(TACGTA) (4, 6). Constitutive nuclear-localized HY5 plays animportant role in integrating light signaling and several phyto-hormone and abiotic stress-related signaling pathways (7). Forexample, during seed germination and early seedling develop-ment, HY5 directly binds to G-box present in the promoter ofABA Insensitive 5 (ABI5) and activates its expression (8). Light andreactive oxygen species (ROS) signaling regulate deetiolation ofseedlings, which is mediated by transcriptional modules regulated bydirect binding of HY5 and phytochrome interacting factor (PIF) TFsto the G-box element present in the promoter of ROS-responsive

genes (9). The cold acclimation response in Arabidopsis is positivelyregulated by HY5 through direct binding to Z-box and other cis-acting elements present in cold-inducible genes (10). These studiessuggest that better performance of plant growth requires the properintegration of light and defense response against abiotic stressors,which is mediated through biding of HY5 to common cis-actingelements. Recent studies have reported that abiotic stresses, suchas heat/cold shock, salt stress, oxidative stress, and osmotic pres-sure, can perturb endoplasmic reticulum (ER) homeostasis,causing the accumulation of misfolded proteins in ER and evokingER stress in plants (11–14). These observations lead us to thinkabout the role of light signals in ER stress response.To mitigate ER stress, the level of protein-folding chaperones

and ER-associated protein degradation (ERAD) machinery isincreased, which is known as unfolded protein response (UPR)(15). Different membrane-associated TFs (MTFs) transducestress signals to the nucleus. Typically, two MTFs, bZIP28 andbZIP60, play vital roles in ensuring cell survival during ER stressin plants (16, 17). These MTFs up-regulate genes encodingcomponents of the ER protein-folding machinery, including lu-minal binding protein (BIP), calnexin (CNX), calreticulin (CRT),and protein disulfide isomerase (PDI). These genes share a con-sensus element in their promoters known as an ER stress responseelement (ERSE), which comprises two subelements: a CACGsubelement that binds bZIP dimers and a CCAAT subelement

Significance

In nature, plants are inevitably exposed to adverse conditionssuch as salinity, drought, and extreme temperatures. Recentreports suggest that environmental stresses critically affect theprotein-folding capacity of endoplasmic reticula (ER), leading toER stress. As the growth and development of plants signifi-cantly depend on light environment, the crosstalk betweenlight signaling and ER stress response explained in current re-search can be a unique feature of plants. Our results suggestthat light increases the ER stress sensitivity of plants andELONGATED HYPOCOTYL 5, a positive regulator of light sig-naling, negatively regulates unfolded protein response geneexpression in plant cells, which decreases the protein-foldingcapacity. The present study may form the basis for designingnew strategies to increase stress tolerance of plants by tightlycontrolling light environment.

Author contributions: C.H.K. and S.Y.L. designed research; G.M.N., P.M., J.H.P., Y.J.J., H.B.C.,Y.H.C., and I.J.J. performed research; W.Y.K. and D.-J.Y. contributed new reagents/analytic tools; G.M.N., C.H.K., and P.M. analyzed data; and C.H.K. and S.Y.L. wrotethe paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

Freely available online through the PNAS open access option.1G.M.N., C.H.K., and P.M. contributed equally to this work.2To whom correspondence should be addressed. Email: sylee@gnu.ac.kr.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1609844114/-/DCSupplemental.

2084–2089 | PNAS | February 21, 2017 | vol. 114 | no. 8 www.pnas.org/cgi/doi/10.1073/pnas.1609844114

Dow

nloa

ded

by g

uest

on

Mar

ch 2

7, 2

021

that binds CCAAT box-binding factors (17, 18). However, pro-teins occasionally fail to mature in the ER and are exported fromthe ER for degradation by the ubiquitin–proteasome system aspart of the ERAD pathway (15, 19). When plants are subjectedto environmental stresses, the levels of such malformed proteinscan exceed the capacity of the ER quality control and ERADsystems and leads to autophagy and programmed cell death toeliminate damaged cells (11, 20–23).Accumulating evidence suggests that different abiotic stresses

induce UPR and that the stress response is integrated with lightsignaling. Moreover, it has recently been suggested that lightenhances UPR through the activation of BIP2 expression; how-ever, the molecular mechanism underlying this process remainsunclear (24). In this study, we suggest that light critically enhancesthe ER stress sensitivity and that a key component of light sig-naling, HY5, acts as a negative regulator of the UPR signalingpathway.

ResultsThe ER Stress Response in Plants Is Critically Affected by LightIntensity. To investigate the effect of light on the ER stress re-sponse, we germinated wild-type (WT) Arabidopsis seeds incontinuous darkness or white light on a medium containingdifferent tunicamycin (Tm) concentrations. Plants grown in thedark exhibited skotomorphogenic development and plant growthwas unaffected by Tm treatment. In contrast, plants grown underwhite light exhibited photomorphogenic development, and plantgrowth was seriously affected by Tm treatment. The stronger thelight intensity, the more severe was the growth inhibition causedby Tm addition. For example, compared with seedlings grownunder 20 μmol m−2·s−1 light, those grown under 100 μmol m−2·s−1

light showed greater Tm sensitivity (Fig. 1A). A significant re-duction of the relative fresh weight was observed with increasedlight intensity in presence of Tm (Fig. 1B). We examined theinhibitory effects of Tm and DTT on primary root elongation.Severe growth inhibition was evident in presence of Tm and DTTwith increasing light intensity (Fig. 1 C and D and SI Appendix,Fig. S1). Thus, the combination of ER stress and light synergis-tically inhibited root growth, demonstrating the significant role oflight in enhancing the ER stress sensitivity. Interestingly, the freshweight of seedlings showed significant recovery in presence of0.2 mM tauroursodexycholic acid (TUDCA), a chemical chap-erone under the Tm-treated condition at 100 μmol m−2·s−1 lightcondition, implying that UPR relieves the Tm sensitivity (SIAppendix, Fig. S2). To confirm roles of UPR, we compared ex-pression of UPR marker genes such as BIP3, CRT1, and CNX1with or without Tm treatment under different light intensities.We found that expression of the UPR marker genes was inducedunder the Tm-treated condition. Moreover, the UPR gene ex-pression was higher in the dark condition, whereas it decreases byincreasing the light intensity (SI Appendix, Fig. S3).

HY5 Positively Mediates the ER Stress Sensitivity. Recent genomicstudies have shown that light induces massive reprogramming ofthe plant transcriptome and that this process is primarily regu-lated by one key TF, HY5 (5). Thus, to determine the involve-ment of HY5 in UPR signaling, an hy5 mutant, an HY5overexpression line P35S:yellow fluorescent protein (YFP or Y)-HY5/hy5, and WT plants were germinated under long-day con-ditions on a medium supplemented with (15 or 30 ng/mL) orwithout (control) Tm. Under control conditions, hy5 mutantsexhibited an elongated hypocotyl, whereas P35S:Y-HY5/hy5 plantsshowed hypocotyl growth similar to that of WT and lacked aninhibitory phenotype (SI Appendix, Fig. S4). However, when thegrowth medium was supplemented with Tm, the effect was clear(Fig. 2 A–C). Under stress conditions, hy5 mutant growth wasless affected, evident in the larger proportion of greenish-big(G-B) plants and smaller proportion of yellowish-small (Y-S) plants

compared with WT plants. In contrast, P35S:Y-HY5/hy5 plants weremore sensitive to Tm-induced ER stress than WT and hy5 plants.We also obtained another hy5mutant, hy5-1 (Landsberg erecta; Ler)and c-hy5, a complementation line of the mutant withHY5 under itsown promoter (25). When we compared phenotypes of the wildtype (Ler), hy5-1, and c-hy5 plants grown under different lightintensity conditions (20 or 150 μmol m−2·s−1) with or without Tmtreatment, hy5-1 was strongly tolerant of ER stress and the tol-erance level was reverted in c-hy5 to a level similar to Ler (SI Ap-pendix, Fig. S5). Consistently, our electrolyte leakage assay revealedthat Tm-induced cell death was enhanced in P35S:Y-HY5/hy5 plants,but reduced in hy5 mutants, compared with that in WT plants(Fig. 2D). Taken together, these findings suggest that hy5 muta-tion conferred ER stress tolerance, whereas HY5 overexpressionenhanced ER stress sensitivity in Arabidopsis.The tolerant phenotype of the hy5 mutant and sensitivity of

P35S:Y-HY5/hy5 plants prompted us to investigate whether HY5regulates genes downstream of UPR in plants. We compared theexpression of genes downstream of UPR in WT, hy5, and P35S:Y-HY5/hy5 plants under normal and ER stress conditions. Theexpression of UPR marker genes such as BIP3, CRT1, and PDI10was induced by Tm treatment in WT plants; however, the in-duction of these marker genes occurred to a lesser degree inP35S:Y-HY5/hy5 plants compared with that in WT plants, whereasto a higher degree in hy5 mutant plants compared with that inWT plants (Fig. 3), suggesting that HY5 acted as a negativeregulator for the expression of genes downstream of UPR.

HY5 Binds to the Promoter of the BIP3 Gene. It has been established thatHY5 binds to G-box (CACGTG) in the promoters of light-responsive

Fig. 1. Light enhances the ER stress sensitivity of Arabidopsis. (A) Pheno-types of 2-wk-old WT (Col-0) seedlings grown on MS medium containing 10or 20 ng/mL of Tm or DMSO in continuous darkness (intensity = 0 μmol m−2·s−1)or light (intensity = 20, 100, or 150 μmol m−2·s−1). (B) Relative fresh weight ofplants (mean ± SEM; n = 3) treated as indicated in A. Fresh weights of seed-lings grown in the Tm-untreated conditions were set at 100%. (C) Primary rootlength phenotype of 7-d-old WT seedlings grown on MS medium containing25 or 50 ng/mL of Tm or DMSO in continuous light (intensity = 20, 100, or150 μmol m−2·s−1). Root lengths of seedlings grown in the Tm-untreatedconditions were set at 100%. (White scale bar, 20 mm.) (D) Relative rootlength of plants (mean ± SEM; n ≥ 20) treated as indicated in C. (B and D)Statistics by t test are shown; *P < 0.05, **P < 0.01, ***P < 0.001, ****P <0.0001, and NS, no significance.

Nawkar et al. PNAS | February 21, 2017 | vol. 114 | no. 8 | 2085

PLANTBIOLO

GY

Dow

nloa

ded

by g

uest

on

Mar

ch 2

7, 2

021

genes (4). We examined the promoter sequence of the BIP3gene and found that it contains two G-box–like sequences withinthe ERSE1 and ERSE2 motifs (SI Appendix, Fig. S6). Thus, totest the hypothesis that HY5 directly binds to the two ERSEmotifs of the BIP3 promoter, we conducted in vitro electro-phoretic mobility shift assays (EMSAs). For this experiment, theG-box of the RBCS1A promoter was used as the positive control(6). Probes containing ERSE1 and ERSE2 from the BIP3 pro-moter and the G-box of the RBCS1A promoter were designedand labeled with biotin (SI Appendix, Table S1 and Fig. S6).Recombinant proteins, maltose-binding protein (MBP) alone andMBP fused to HY5 (MBP-HY5), produced in Escherichia coliwere incubated with the biotin-labeled probes. MBP-HY5, but notMBP alone, caused a mobility shift of the three probes (Fig. 4A).Furthermore, the addition of unlabeled ERSE competitors di-minished the intensities of the shifted bands, indicating that HY5bound specifically to both ERSE motifs and the control G-box(Fig. 4B).To confirm the binding of HY5 to particular ERSE sequences

in vivo, we performed ChIP experiments using P35S:Y-HY5/hy5seedlings. After the immunoprecipitation of protein–DNA com-plexes using an anti-GFP antibody, the DNA fragments werequantified by quantitative real-time PCR (qRT-PCR) (Fig. 4C).Primers for the TA3 promoter were used as the negative control(26). The occupancy of HY5 was remarkably high in the “P1”and “P2” regions, which contain ERSE1 and ERSE2 of the BIP3promoter, respectively, compared with that in negative controlsite in P35S:Y-HY5/hy5 plants but not in the hy5 mutants (Fig.4D). These results confirm that HY5 directly associated with the

ERSE motifs of the BIP3 promoter in vitro and in vivo and thatthe binding of HY5 to the BIP3 promoter was reduced under theER stress. To know the reason behind reduction of the bind-ing affinity of HY5 in ER stress compared with normal condi-tions, we investigated the levels of HY5 protein in 10-d-oldP35S:Y-HY5/hy5 seedlings in the presence or absence of Tmwith or without supplementation of the proteasome inhibitorMG132 for 12 h. The total proteins were extracted and subjectedto immunoblotting analysis (Fig. 4E). Under control light condi-tions, HY5 protein level was relatively stable, showing no signifi-cant difference. The HY5 protein was clearly degraded by Tmtreatment, and this Tm-induced degradation of HY5 was signifi-cantly inhibited by MG132 treatment (Fig. 4F), indicating that theHY5 protein was subjected to 26S proteasome-mediated deg-radation under ER stress conditions.

HY5 Competes with bZIP28 to Bind to the ERSE Motifs in the BIP3Promoter. Previous studies have shown that the core CACG ele-ment of the ERSE motif is important for bZIP28 binding, whereasthe CCAAT element is critical for nuclear factor Y (NF-Y)binding (SI Appendix, Fig. S7) (17, 18). Because HY5 binding tothe BIP3 promoter was reduced under ER stress conditions (Fig.4D), we tested the possibility that bZIP28 exerts a negative effecton HY5 binding to the ERSE motifs. To confirm this hypothesis,we focused on the binding sites in the ERSE motif. We performedan EMSA assay with various point-mutated DNA probes con-taining ERSE1 and ERSE2 motifs of the BIP3 promoter (SIAppendix, Table S1 and Fig. S7). HY5 strongly bound to the WTERSE1 and ERSE2 motifs (Fig. 5A; lane 2, both panels); how-ever, nucleotide base substitutions in the bZIP binding elementsof the ERSE motifs significantly prevented mobility shifting(lane 4, both panels). In contrast, nucleotide base substitution inthe NF-Y binding sites (lane 3, both panels) and nucleotide baseaddition or deletion (lanes 5 and 6, both panels) had no effect onmobility shifting, indicating that HY5 bound to the bZIP28binding site but not to the NF-Y binding sites.Using the EMSA assay and ERSE1 probe, we tested whether

HY5 and bZIP28 compete with each other to bind to the ERSEmotifs (Fig. 5B). The mobility of the probe was shifted in aconcentration-dependent manner by MBP-HY5 and MBP-bZIP28(lanes 3–8) but was unaffected by MBP alone (lanes 1–2). Whenthe amount of MBP-bZIP28 was fixed and increasing amounts ofMBP-HY5 were added, the formation of the MBP-bZIP28-probecomplex was dramatically decreased, whereas that of the MBP-HY5-probe complex increased in a concentration-dependentmanner (lanes 9–11). Moreover, MBP-bZIP28 was able to re-place MBP-HY5 for the formation of the complex with probe (SIAppendix, Fig. S8). However, the addition of MBP alone did notaffect the formation of the MBP-bZIP28-probe complex (com-pare lanes 8 and 12 in Fig. 5B). These suggest that bZIP28 andHY5 specifically competed to bind to the ERSE motifs. In

Fig. 2. HY5 positively mediates the ER stress sensitivity. (A) Phenotypes of 2-wk-old WT, hy5, and P35S:Y-HY5/hy5 seedlings grown on MS medium con-taining 15 or 30 ng/mL of Tm or DMSO under LD (16-h light/8-h dark)conditions. (B) Percentage of G-B, G-S, and Y-S plants (mean ± SEM; n = 3)treated as indicated in A. (C) Relative fresh weight of plants (mean ± SEM;n = 3) treated as indicated in A. Fresh weights of seedlings grown in the Tm-untreated conditions were set at 100%. (D) WT, hy5, and P35S:Y-HY5/hy5seedlings were grown on MS medium for 2 wk and treated with or without5 μg/mL of Tm for 6 h. They were used for ion leakage (mean ± SEM; n = 3)after 24-h recovery. (B–D) Statistics by t test are shown; *P < 0.05, **P < 0.01,***P < 0.001, ****P < 0.0001, and NS, no significance.

Fig. 3. HY5 suppresses the expression of UPR marker genes. qRT-PCR analysisof transcript levels of BIP3, CRT1, and PDI10 in WT, hy5, and P35S:Y-HY5/hy5plants at the indicated time points after Tm treatment. Error bars denote ±SEM.

2086 | www.pnas.org/cgi/doi/10.1073/pnas.1609844114 Nawkar et al.

Dow

nloa

ded

by g

uest

on

Mar

ch 2

7, 2

021

conclusion, HY5 and bZIP28 shared binding sites in the BIP3promoter and competed to bind to the ERSE motifs in vitro.We performed a transient transcription assay in tobacco leaves

to investigate whether HY5 functions in the regulation of BIP3gene expression in vivo. We constructed luciferase (LUC) re-porter plasmids in which LUC reporter-gene expression wasdriven by the BIP3 promoter containing the ERSE motifs (Fig.5C). Increased expression of the reporter gene in response to Tmtreatment was confirmed to validate the functionality of theconstruct. As shown in Fig. 5D, transiently expressed bZIP28activated BIP3 expression, consistent with previous reports (18).HY5 alone, however, had no effect on BIP3 expression, regardlessof Tm treatment. Interestingly, the coexpression of HY5 andbZIP28 dramatically repressed bZIP28-activated reporter geneexpression, confirming the negative role of HY5 on bZIP28-driven transcriptional activation.To confirm the antagonistic relationship between HY5 and

bZIP28, we created a hy5 bzip28 double mutant by crossing hy5

and bzip28 single mutants (SI Appendix, Fig. S9) and analyzedthe genetic interaction between HY5 and bZIP28 for UPR. Weexamined expression of UPR marker genes such as BIP3 andPDI10 in WT, hy5, bzip28, and hy5 bzip28 plants over the courseof Tm treatment (SI Appendix, Fig. S9). qRT-PCR showed thatthe UPR gene expression was lower in bzip28 and hy5 bzip28mutants than in WT plants; moreover, bzip28 and hy5 bzip28plants exhibited similar levels of the expression, implying that thenegative roles of HY5 in the UPR were mostly bZIP28 de-pendent. Furthermore, the hy5 mutant showed a much higherinduction of the UPR genes compared to the other plants. Thesefindings suggest that HY5 negatively regulates the UPR geneexpression and antagonistically acts on the bZIP28-mediated geneactivation.

DiscussionLight is one of the main factors required for proper plant growthand development, from seed germination to flowering and seedmaturation (2, 27). Recently, crosstalk in plants between lightsignaling and stress responses has become a hot research topic.However, to date, the influence of light on the ER stress re-sponse has not been studied. To analyze the involvement of lightin the ER stress response in plants, we investigated the effect ofER stress in WT Arabidopsis in the presence or absence of light.We found that light is required for the ER stress-induced growthinhibition of plants and that increasing light intensity enhances

Fig. 5. HY5 competes with bZIP28 to bind to the ERSE motifs of the BIP3promoter. (A) EMSA of the binding of MBP-HY5 to the core CACG subele-ment (bZIP28 binding site) present in ERSE motifs of the BIP3 promoter.Different ERSE probes (WT, M1–M4 shown in SI Appendix, Fig. S7) were al-tered by multiple substitutions (M1 and M2), or the length of spacers (M3and M4) was altered. The black arrowhead shows the free probe. (B) EMSAshowing the competition between MBP-HY5 and MBP-bZIP28ΔC to bind tothe ERSE motifs of the BIP3 promoter. Superscript numbers 1X–3X representthe increasing protein concentration. The black arrowhead shows the freeprobe. (C) Schematic representation of constructs used in the transienttranscription assay in tobacco leaves. A reporter vector PBIP3:LUC containingthe BIP3 promoter (942 bp upstream of the start codon) driving LUC andP35S:GUS was used as an internal control. P35S:Y-HY5 and P35S:HA-bZIP28were used as effector constructs. (D) Activation of PBIP3:LUC by differentcombinations of effectors. (Lower) Superscript numbers indicate the ratioof OD600 of Agrobacterium used in coinfiltration. The activation value ofPBIP3:LUC under control conditions without any effectors was set at 1 (mean± SEM; n = 3).

Fig. 4. HY5 binds to the promoter of BIP3 gene and undergoes 26S pro-teasome-mediated degradation in ER stress. (A) EMSA of the binding of therecombinant MBP-HY5 to ERSE1 and ERSE2 motifs of the BIP3 promoter andG-box motif of the RBCS1A promoter (positive control) in a concentration-dependent manner. (B) Respective unlabeled probes (with the same se-quence as the biotin-labeled probes) were used as competitors. (A and B)SuperScript numbers, 1X–3X represent the increasing concentration of un-labeled probes. (C) Schematic representation of the BIP3 promoter with thelocation of the two ERSE motifs. P1 and P2 represent the respective primerpositions used for ChIP-qPCR. (D) ChIP-qPCR analyses of HY5 binding to theBIP3 promoter. ChIP assays were performed on hy5 and P35S:Y-HY5/hy5seedlings treated with Tm (+Tm) or DMSO (−Tm) for 12 h. DNA–proteincomplexes were immunoprecipitated using antibodies against GFP and rabbitIgG (negative control). ChIP DNA was quantified by qRT-PCR with primersspecific to the ERSE motifs (P1 and P2) and TA3 promoter (control; mean ±SEM; n = 3 technical replicates). (E) Immunoblotting analysis of P35S:Y-HY5/hy5 seedlings expressing YFP-HY5 treated with 5 μg/mL of Tmwith or without50 μM of MG132 for the indicated time. (Upper) Representative Western blot.(Lower) Ponceau S staining of same blot, which served as a loading control.(F) Quantification of relative abundance (fold) of YFP-HY5 protein levelscompared with control conditions. The data are shown as means ± SEM fromthree independent biological repeats. (B and D) Statistics by t test are shown;*P < 0.05, **P < 0.01, ***P < 0.001, and NS, no significance.

Nawkar et al. PNAS | February 21, 2017 | vol. 114 | no. 8 | 2087

PLANTBIOLO

GY

Dow

nloa

ded

by g

uest

on

Mar

ch 2

7, 2

021

the sensitivity of plants to ER stress. Based on these results, weattempted to identify the molecular links between light signalingand UPR.In this study, we found that a positive regulator of light sig-

naling, HY5, negatively regulates UPR. The loss of function ofHY5 rendered plants tolerant to ER stress induced by Tm treat-ment. Previously, it has been established that ER stress evoked byenvironmental conditions activates the two arms of UPR and∼0.7% of the genome is up-regulated as a part of UPR in Ara-bidopsis (28). The activity of bZIP28 and bZIP60 is required forthe induction of UPR and stress-tolerance phenotype in plants(29, 30). Because HY5 is a well-known TF, we hypothesized thatit plays an important role in UPR by regulating gene expression.In fact, our qRT-PCR data suggest that the tolerance of thehy5 mutant to ER stress conditions was due to the elevated ex-pression of UPR marker genes in these plants. In conclusion,these results are consistent with those of recent studies sug-gesting that HY5 is a critical component in the integration oflight signaling with different defense responses against stressconditions, an important event for plant survival under adverseconditions (10, 31, 32).As shown by our EMSA results, HY5 specifically bound to the

bZIP28 binding sites of ERSE1 and ERSE2 in the BIP3 pro-moter. Furthermore, the binding of bZIP28 to the ERSE motifwas reduced when the HY5 protein was present in excess andvice versa. This analysis illustrates the competition between HY5and bZIP28 for binding to the BIP3 promoter, emphasizing theimportance of the concentrations of particular TFs. This idea isalso supported by our transient transcription assay in tobacco,which showed that HY5 inhibited transcriptional activation bybZIP28. Previously, it has been reported that bZIP28 forms atranscriptional complex with NF-Y TFs, which is important forfull induction of the BIP3 gene (18). Recently, bZIP28 has beenshown to be involved in formation of a transcriptional complexwith COMPASS-like components, which are involved in histoneH3K4 trimethylation (H3K4me3), which correlates with activegene expression (33). Moreover, it is also known that histonemodification plays an important role in HY5-mediated geneexpression (4, 34). Thus, it is likely to form such transcriptionalcomplexes with bZIP28 under ER stress conditions, which caninterfere with HY5 binding to the ERSE motifs. Our ChIP assayalso supports this notion, as the binding of HY5 is decreasedunder ER stress conditions (Fig. 4). Our results imply that HY5exerted an antagonistic effect on bZIP28, but we cannot excludethe possibility that HY5 affects activity of other TFs, such asbZIP60, NAC062, and NAC103, which regulate the expressionof UPR marker genes required for ER stress tolerance (35–38).Our data suggest that HY5 protein level is regulated in ER

stress conditions through the 26S proteasomal degradation sys-tem, which is similar to light/dark regulation of the HY5 levelduring the deetiolation process (3). Previous studies also suggestthat the abundance of HY5 protein is regulated by either lowtemperature or short heat shock through the 26S proteasome-mediated degradation system (10, 31). The proteosomal degra-dation of HY5 and proteolytic cleavage of bZIP28 fine tunes thebalance between the two proteins to compete for binding theERSE motifs and in turn regulate UPR genes. Nearly 50% deg-radation of HY5 protein suggests that it may have other functionsthat should be identified in future studies.Considering these results together, it is possible to propose a

hypothetical model that explains the roles of HY5 and bZIP28 inmediating crosstalk between light signaling and UPR (Fig. 6).Under normal light conditions, HY5 binds to not only G-box el-ements present in the promoter of light-response genes to regu-late photomorphogenesis, but also to the ERSE motifs present inthe promoters of UPR genes to reduce their expression to basallevels. Under these conditions, bZIP28 is anchored to the ERmembrane through its transmembrane domain. Under ER stress,

bZIP28 is activated by proteolysis and the activated bZIP28ΔCtranslocates to the nucleus, where it competes with HY5 to bind tothe ERSE motifs present in the promoters of UPR genes. Underthese conditions, active bZIP28ΔC accumulates in the nucleus,whereas HY5 is removed by the 26S proteasome system, resultingin the binding of bZIP28ΔC to the ERSE motifs and the up-reg-ulation of UPR genes. This study may contribute to a better un-derstanding of crosstalk between light signaling and UPR in plants.

Materials and MethodsPlant Materials and Growth Conditions. Arabidopsis WT, T-DNA mutants, andtransgenic lines were prepared in the Columbia (Col-0) ecotype background.We obtained the homozygous T-DNA lines hy5 (SALK_096651C) and bzip28(SALK_132285) from the Arabidopsis Biological Resource Center, The OhioState University, Columbus, OH. Double-mutant hy5 bzip28 was generatedby genetic crossing using hy5 female and bzip28 male parents, and homo-zygous lines were confirmed by PCR genotyping (SI Appendix, Fig. S9). Weobtained mutant seeds hy5-1 and PHY5:HY5-Y/hy5-1 (c-hy5) in Landsbergerecta (Ler) background as generous gifts from R. Ulm, University of Geneva,Geneva, Switzerland (25). Seedlings were grown on half-strength Murashigeand Skoog (MS) medium containing 2% (wt/vol) sucrose and 0.25% (wt/vol)Phytagel (Sigma-Aldrich), pH 5.7. Details on growth conditions, stresstreatments for seedlings, and other methods used in this study are describedin SI Appendix, SI Materials and Methods.

Construction of Transgenic Plants. All DNA constructs and transgenic plantsprepared in this study are described in SI Appendix, SI Materials andMethods.

Measurement of Electrolyte Leakage. Details are in SI Appendix, SI Materialsand Methods.

Fig. 6. Schematic model showing HY5-mediated crosstalk between lightsignaling and UPR. Under normal light condition, HY5 positively regulateslight-responsive genes and directly binds to the ERSE motifs present in thepromoter of UPR genes to repress their expression. At the same time, bZIP28is anchored to the ER membrane in a dormant form. Under ER stress con-ditions, bZIP28 is translocated to the Golgi apparatus, where it is pro-teolytically processed, after which it translocates to the nucleus, where itcompetes with HY5 to bind to the ERSE motifs. The HY5 protein is subjectedto 26S proteasomal degradation, resulting in higher levels of active bZIP28 inthe nucleus, which turn on UPR for plant survival.

2088 | www.pnas.org/cgi/doi/10.1073/pnas.1609844114 Nawkar et al.

Dow

nloa

ded

by g

uest

on

Mar

ch 2

7, 2

021

EMSA and ChIP Assay. Details are in SI Appendix, SI Materials and Methods.All primer sequences are listed in SI Appendix, Table S2.

Immunoblotting Analysis. Details are in SI Appendix, SI Materials and Methods.

Transient Transcription Assay.Details are in SI Appendix, SI Materials andMethods.

Total RNA Isolation and Quantitative Real-Time PCR Analysis. Details are in SIAppendix, SI Materials and Methods. All primer sequences are listed in SIAppendix, Table S2.

Statistical Analysis. Statistical significances were determined using Student’s ttest. P values were calculated using GraphPad QuickCalcs (available online atwww.GraphPad.com/).

ACKNOWLEDGMENTS. We thank Dr. R. Ulm and C. Richard (University ofGeneva) for their kind provision of hy5-1 and PHY5:HY5-Y/hy5-1 (c-hy5) seeds.This work was supported by the Next-Generation BioGreen 21 program (Sys-tems and Synthetic Agrobiotech Center, Grant PJ011379), Rural Develop-ment Administration, and the Basic Science Research Program through theNational Research Foundation of Korea funded by the Ministry of Education(Grant 2016R1D1A1B01016551), Republic of Korea.

1. Rizzini L, et al. (2011) Perception of UV-B by the Arabidopsis UVR8 protein. Science332(6025):103–106.

2. Jiao Y, Lau OS, Deng XW (2007) Light-regulated transcriptional networks in higherplants. Nat Rev Genet 8(3):217–230.

3. Osterlund MT, Hardtke CS, Wei N, Deng XW (2000) Targeted destabilization of HY5during light-regulated development of Arabidopsis. Nature 405(6785):462–466.

4. Zhang H, et al. (2011) Genome-wide mapping of the HY5-mediated gene networks inArabidopsis that involve both transcriptional and post-transcriptional regulation.Plant J 65(3):346–358.

5. Lee J, et al. (2007) Analysis of transcription factor HY5 genomic binding sites revealedits hierarchical role in light regulation of development. Plant Cell 19(3):731–749.

6. Chattopadhyay S, Ang LH, Puente P, Deng XW, Wei N (1998) Arabidopsis bZIP proteinHY5 directly interacts with light-responsive promoters in mediating light control ofgene expression. Plant Cell 10(5):673–683.

7. Lau OS, Deng XW (2010) Plant hormone signaling lightens up: Integrators of light andhormones. Curr Opin Plant Biol 13(5):571–577.

8. Chen H, et al. (2008) Integration of light and abscisic acid signaling during seedgermination and early seedling development. Proc Natl Acad Sci USA 105(11):4495–4500.

9. Chen D, et al. (2013) Antagonistic basic helix-loop-helix/bZIP transcription factorsform transcriptional modules that integrate light and reactive oxygen species sig-naling in Arabidopsis. Plant Cell 25(5):1657–1673.

10. Catalá R, Medina J, Salinas J (2011) Integration of low temperature and light signalingduring cold acclimation response in Arabidopsis. Proc Natl Acad Sci USA 108(39):16475–16480.

11. Faria JA, et al. (2011) The NAC domain-containing protein, GmNAC6, is a downstreamcomponent of the ER stress- and osmotic stress-induced NRP-mediated cell-deathsignaling pathway. BMC Plant Biol 11:129.

12. Gao H, Brandizzi F, Benning C, Larkin RM (2008) A membrane-tethered transcriptionfactor defines a branch of the heat stress response in Arabidopsis thaliana. Proc NatlAcad Sci USA 105(42):16398–16403.

13. Liu JX, Srivastava R, Che P, Howell SH (2007) Salt stress responses in Arabidopsis utilizea signal transduction pathway related to endoplasmic reticulum stress signaling. PlantJ 51(5):897–909.

14. Williams B, Kabbage M, Britt R, Dickman MB (2010) AtBAG7, an Arabidopsis Bcl-2-associated athanogene, resides in the endoplasmic reticulum and is involved in theunfolded protein response. Proc Natl Acad Sci USA 107(13):6088–6093.

15. Liu JX, Howell SH (2010) Endoplasmic reticulum protein quality control and its re-lationship to environmental stress responses in plants. Plant Cell 22(9):2930–2942.

16. Iwata Y, Koizumi N (2012) Plant transducers of the endoplasmic reticulum unfoldedprotein response. Trends Plant Sci 17(12):720–727.

17. Howell SH (2013) Endoplasmic reticulum stress responses in plants. Annu Rev PlantBiol 64:477–499.

18. Liu JX, Howell SH (2010) bZIP28 and NF-Y transcription factors are activated by ERstress and assemble into a transcriptional complex to regulate stress response genes inArabidopsis. Plant Cell 22(3):782–796.

19. Hüttner S, Strasser R (2012) Endoplasmic reticulum-associated degradation of glyco-proteins in plants. Front Plant Sci 3:67.

20. Yang ZT, et al. (2014) The membrane-associated transcription factor NAC089 controlsER-stress-induced programmed cell death in plants. PLoS Genet 10(3):e1004243.

21. Watanabe N, Lam E (2008) BAX inhibitor-1 modulates endoplasmic reticulum stress-mediated programmed cell death in Arabidopsis. J Biol Chem 283(6):3200–3210.

22. Mendes GC, et al. (2013) GmNAC30 and GmNAC81 integrate the endoplasmic re-ticulum stress- and osmotic stress-induced cell death responses through a vacuolarprocessing enzyme. Proc Natl Acad Sci USA 110(48):19627–19632.

23. Liu Y, et al. (2012) Degradation of the endoplasmic reticulum by autophagy duringendoplasmic reticulum stress in Arabidopsis. Plant Cell 24(11):4635–4651.

24. Lu DP, Christopher DA (2008) Light enhances the unfolded protein response asmeasured by BiP2 gene expression and the secretory GFP-2SC marker in Arabidopsis.Physiol Plant 134(2):360–368.

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

26. Yamaguchi N, et al. (2014) PROTOCOLS: Chromatin immunoprecipitation from Ara-bidopsis tissues. Arabidopsis Book 12:e0170.

27. Kami C, Lorrain S, Hornitschek P, Fankhauser C (2010) Light-regulated plant growthand development. Curr Top Dev Biol 91:29–66.

28. Martínez IM, Chrispeels MJ (2003) Genomic analysis of the unfolded protein responsein Arabidopsis shows its connection to important cellular processes. Plant Cell 15(2):561–576.

29. Sun L, et al. (2013) The lumen-facing domain is important for the biological functionand organelle-to-organelle movement of bZIP28 during ER stress in Arabidopsis. MolPlant 6(5):1605–1615.

30. Deng Y, Srivastava R, Howell SH (2013) Protein kinase and ribonuclease domains ofIRE1 confer stress tolerance, vegetative growth, and reproductive development inArabidopsis. Proc Natl Acad Sci USA 110(48):19633–19638.

31. Karayekov E, Sellaro R, Legris M, Yanovsky MJ, Casal JJ (2013) Heat shock-inducedfluctuations in clock and light signaling enhance phytochrome B-mediated Arabi-dopsis deetiolation. Plant Cell 25(8):2892–2906.

32. Xu D, et al. (2014) Convergence of light and ABA signaling on the ABI5 promoter.PLoS Genet 10(2):e1004197.

33. Song ZT, et al. (2015) Transcription factor interaction with COMPASS-like complexregulates histone H3K4 trimethylation for specific gene expression in plants. Proc NatlAcad Sci USA 112(9):2900–2905.

34. Wang Y, Wang Y, Song Z, Zhang H (2016) Repression of MYBL2 by both microRNA858aand HY5 leads to the activation of anthocyanin biosynthetic pathway in Arabidopsis.Mol Plant 9(10):1395–1405.

35. Yang ZT, et al. (2014) A plasma membrane-tethered transcription factor, NAC062/ANAC062/NTL6, mediates the unfolded protein response in Arabidopsis. Plant J 79(6):1033–1043.

36. Tajima H, Iwata Y, Iwano M, Takayama S, Koizumi N (2008) Identification of anArabidopsis transmembrane bZIP transcription factor involved in the endoplasmicreticulum stress response. Biochem Biophys Res Commun 374(2):242–247.

37. Sun L, et al. (2013) The plant-specific transcription factor gene NAC103 is induced bybZIP60 through a new cis-regulatory element to modulate the unfolded protein re-sponse in Arabidopsis. Plant J 76(2):274–286.

38. Iwata Y, Koizumi N (2005) An Arabidopsis transcription factor, AtbZIP60, regulatesthe endoplasmic reticulum stress response in a manner unique to plants. Proc NatlAcad Sci USA 102(14):5280–5285.

Nawkar et al. PNAS | February 21, 2017 | vol. 114 | no. 8 | 2089

PLANTBIOLO

GY

Dow

nloa

ded

by g

uest

on

Mar

ch 2

7, 2

021