lable at ScienceDirect

Plant Physiology and Biochemistry 74 (2014) 176e184

Contents lists avai

Plant Physiology and Biochemistry

journal homepage: www.elsevier .com/locate/plaphy

Research article

AtObgC-AtRSH1 interaction may play a vital role in stress responsesignal transduction in Arabidopsis

Ji Chen a,b,1, Woo Young Bang c,1, Yuno Lee b, Songmi Kim b, Keun Woo Lee b, Se Won Kimd,Young Sim Son b, Dae Won Kimb, Salina Akhter b, Jeong Dong Bahk b,*

aAgronomy College, Sichuan Agricultural University, Chengdu 611130, ChinabDivision of Applied Life Sciences (BK21þ), Graduate School of Gyeongsang National University, Jinju 660-701, Republic of KoreacDepartment of Horticultural Sciences, Texas A&M University, College Station, TX 77843-2133, USAdGreen Bio Research Center, Cabbage Genomics Assisted Breeding Supporting Center, Korea Research Institute of Bioscience and Biotechnology (KRIBB),Daejeon 305-806, Korea

a r t i c l e i n f o

Article history:Received 26 May 2013Accepted 16 October 2013Available online 22 November 2013

Keywords:ChloroplastAtObgCAtRSH1ppGppStress responseArabidopsis thaliana

Abbreviations: ACT, aspartate kinase-chorismatethaliana; BiFC, bimolecular fluorescence complemdomain; DS 3.0 software, discovery studio version 3.potentials; MEGA 5.05, molecular evolutionary genetdomain; Obg, Spo0B-associated GTP-binding proteinObg GTPase; ONPG, o-nitrophenyl-b-D-galactopyranobispyrophosphate; qRT-PCR, quantitative real-timedesiccation 29A; RSH, RelA/SpoT homolog; RT-PCR, rethreonyl-tRNA synthetase-GTPase-SpoT proteins; TP, tWaals; VSP2, vegetative storage protein 2.* Corresponding author. Tel.: þ82 55 772 2607; fax

E-mail address: jdbahk@gnu.ac.kr (J.D. Bahk).1 These authors contributed equally to this work.

0981-9428/$ e see front matter � 2013 Elsevier Mashttp://dx.doi.org/10.1016/j.plaphy.2013.10.022

a b s t r a c t

The interaction of Obg (Spo0B-associated GTP-binding protein) GTPase and SpoT, which is a bifunctionalppGpp (guanosine 30 ,50-bispyrophosphate) hydrolase/synthetase, is vital for the modulation of intra-cellular ppGpp levels during bacterial responses to environmental cues. It has been recently reportedthat the ppGpp level is also inducible by various stresses in the chloroplasts of plant cells. However, thefunction of the ObgeSpoT interaction in plants remains elusive. The results from the present and pre-vious studies suggest that AtRSH1 is a putative bacterial SpoT homolog in Arabidopsis andthat its transcription levels are responsive to wounding and salt stresses. In this study, we used ayeast two-hybrid analysis to map the regions required for the AtObgCeAtRSH1 interaction. Moreover,proteineprotein docking simulations revealed reasonable geometric and electrostatic complementarityin the binding surfaces of the two proteins. The data support our experimental results, which suggestthat the conserved domains in AtObgC and the N terminus of AtRSH1 containing the TGS domaincontribute to their interaction. In addition, quantitative real-time PCR (qRT-PCR) analyses showed thatthe expression of AtObgC and AtRSH1 exhibit a similar inhibition pattern under wounding and salt-stressconditions, but the inhibition pattern was not greatly influenced by the presence or absence of light.Based on in vivo analyses, we further confirmed that the AtRSH1 and AtObgC proteins similarly localize inchloroplasts. Based on these results, we propose that the AtObgCeAtRSH1 interaction plays a vital role inppGpp-mediated stress responses in chloroplasts.

� 2013 Elsevier Masson SAS. All rights reserved.

1. Introduction

The stringent response is a wide-spread bacterial adaptationsystem for global adjustments in response to nutrient limitation

mutase-TyrA; At, Arabidopsisentation; CTD, C-terminal0 software; ESP, electrostaticics analysis; NTD, N-terminal; ObgC, chloroplast-targetingside; ppGpp, guanosine 30 ,50-PCR; RD29A, responsive toverse transcription PCR; TGS,ransit peptide; VDW, van der

: þ82 55 762 2816.

son SAS. All rights reserved.

and adverse environmental stresses. This response is mediated bythe alarmone nucleotide guanosine 30,50-bispyrophosphate(ppGpp) (Potrykus and Cashel, 2008). In Escherichia coli, RelA syn-thesizes ppGpp, and SpoT either synthesizes or degrades ppGpp(Mittenhuber, 2001). However, the SpoT hydrolase activity isessential to bacterial cells because high levels of ppGpp disrupt thecell cycle (Xiao et al., 1991). It was recently reported that the Vibriocholera G protein Obg functions as a repressor of the stringentresponse by interacting with SpoT to maintain low ppGpp levelsunder a nutrient-rich environment (Raskin et al., 2007). Obg(Spo0B-associated GTP-binding protein) (Wout et al., 2004) is amember of the Obg GTPase subfamily, which broadly exists in bothprokaryotes and eukaryotes. In addition, it has been reported thatbacterial Obg homologs (from Caulobacter crescentus, Streptomycescoelicolor, and Vibrio harveyi) are essential for cell growth,morphological differentiation, and DNA replication (Maddock et al.,1997; Okamoto and Ochi, 1998; Slominska et al., 2002).

J. Chen et al. / Plant Physiology and Biochemistry 74 (2014) 176e184 177

ppGpp is found in the chloroplasts of plant cells and is markedlyelevated by plant hormones and biotic and abiotic stresses(Takahashi et al., 2004). Plant RelA/SpoT homolog (RSH) proteinshave also been found in the chloroplasts of Arabidopsis(vanderBiezen et al., 2000), tobacco (Givens et al., 2004), rice(Tozawa et al., 2007), and Chlamydomonas reinhardtii (Kasai et al.,2002). In the model plant Arabidopsis, four RSHs (AtRSH1,AtRSH2, AtRSH3, and AtCRSH) are exclusively nuclear-encodedproteins and are located in the chloroplasts (vanderBiezen et al.,2000; Masuda et al., 2008). These proteins display differentppGpp synthetase activities, organ specificity, and circadianrhythms-dependent expressions (Mizusawa et al., 2008). Further-more, the induction of AtRSH1 and AtRSH2 transcripts is responsiveto wounding and salt stresses (Mizusawa et al., 2008). The detailedmolecular mechanism and additional components of this processare still unknown in Arabidopsis. Nevertheless, our recent findingthat AtObgC strongly interacts with AtRSH1 suggests that AtObgCmay be a novel regulatory factor in ppGpp signaling in chloroplasts(Bang et al., 2012).

In some previous reports, AtObgC, as a chloroplast-targeted Obghomolog GTPase, was found to be indispensable for embryogenesis(Bang et al., 2009; Chigri et al., 2009), plastid ribosome biogenesis(Bang et al., 2012), and thylakoid membrane formation (Garciaet al., 2010) during chloroplast development. The chloroplasts ofalgae and land plants are considered to be originated from ancientcyanobacteria-like prokaryotes. Therefore, chloroplasts conceivablyreserve adaptive stress-responsive functions mediated via ppGppsignaling (Braeken et al., 2006). In this study, we used yeast two-hybrid assays and computational docking simulations to demon-strate that all of the conserved domains of AtObgC (Obg fold, Gdomain, and OCT domain) and the N terminus of AtRSH1 thatcontains the TGS domain are essential for their interaction. We alsofound that the expression of AtObgC and AtRSH1 genes is sup-pressed in a similar manner upon exposure to wounding and saltstress. Taken together, the results obtained in this study suggestthat plant AtObgC may play a significant role in the stress-responsive ppGpp signaling pathway by interacting with AtRSH1in chloroplasts, similarly to bacterial Obgs.

Fig. 1. Dissection of the AtObgC domains that interact with AtRSH1. (A) Functional domain sand AD used in this assay indicate the LexA DNA binding domain and the B42 activation domAtRSH1(78-883) were tested for LacZ activity using liquid ONPG as the b-galactosidase suAtObgC mutants with AtRSH1(78e883), was divided by the activity of the corresponding busing three transformants were performed. The error bars indicate the standard deviations

2. Results and discussion

2.1. Three domains of AtObgC208e681 are required for interactionwith AtRSH178e883

Although RelA and SpoT share high similarity in their amino acidsequences, in E. coli and V. cholera, Obg only interacts with SpoT andnot RelA (Raskin et al., 2007; Wout et al., 2004; Jiang et al., 2007).Based on their primary structures, of the four AtRSHs in Arabidopsis,only AtRSH1, which contains the conserved TGS (threonyl-tRNAsynthetase-GTPase-SpoT proteins) and ACT (aspartate kinase-chorismate mutase-TyrA) domains, is equivalent to bacterial RelA/SpoT (Atkinson et al., 2011). To identify the Arabidopsis homolog ofE. coli SpoT, we performed a phylogenetic analysis using the aminoacid sequences of all four Arabidopsis thaliana AtRSHs and E. coliSpoT and RelA. The phylogenetic analysis also displays that AtRSH1rather than AtRSH2, AtRSH3, and AtCRSH is closer to E. coli SpoT,which further supports the finding that AtRSH1 is an Arabidopsishomolog of the bacterial SpoT protein and may possess specificppGpp hydrolysis activity, similarly to E. coli SpoT (Fig. S1). Thefinding that AtObgC strongly interacts with AtRSH1 and that its OCTdomain is required for this interaction appears to bewell consistentwith our previous report (Bangetal., 2012).

To reveal the role of all of the conserved domains of AtObgC inits interaction with AtRSH178e883, four truncated and three G-domain mutated AtObgC constructs were generated (Fig. 1A).Compared with AtObgC(208e681) (Fig. 1Ba), the truncated con-structs showed significantly lower interaction with AtRSH178e883(Fig. 1Bbee), which indicates that the Obg fold, the G domain, andthe OCT domain of AtObgC are necessary for this interaction. In thisexperiment, 1e206 amino acids of the AtObgC N terminus were notused for the interaction analysis because these consists of a chlo-roplast transit peptide (TP) and a disordered region containingmany glutamate and aspartate residues that induce autonomousactivity. In addition, to examinewhich type of guanosine nucleotidebinding forms among the GTP-binding, GDP-binding, andnucleotide-free forms is appropriate for the interaction, three pointmutants of the G domain were constructed according to a previous

tructure of the full-length AtObgC (at top), as described in ‘Materials and Methods’. BDain, respectively. Yeast strains containing various AtObgCmutant constructs (aeh) withbstrate (B). Each b-galactosidase activity, which represents the interaction of variousait construct with the prey vector only. For reproducibility, at least triplicate analyses(n ¼ 3).

J. Chen et al. / Plant Physiology and Biochemistry 74 (2014) 176e184178

report (OkamotoandOchi, 1998), as shown in Fig. 1Afeh. The GTP-bound AtObgC displayed stronger interaction with AtRSH178e883(Fig. 1Bf) compared with the GDP-bound and nucleotide-free states(Fig. 1Bg and h). Different guanosine binding forms exhibiteddistinct binding activities with AtRSH1, which suggests thatconformational changes of the G domain may affect the interactionbetween AtRSH1 and AtObgC. This finding indicates that the threeconserved domains of AtObgC are required for its interaction withAtRSH1.

2.2. The TGS domain and the N-terminal domain of AtRSH1 arecrucial for interaction with AtObgC

Based on the existence of a TGS homolog domain in both Obgand SpoT protein families (Wolf et al., 1999), it was hypothesizedthat the proteins in both families are involved in ppGpp signalingvia their TGS domains (Battesti and Bouveret, 2006). To definewhich domain(s) of AtRSH1 is(are) responsible for its interactionwith AtObgC, various deletion mutants of AtRSH1were constructedbased on the structural characteristics of the RSH protein (Fig. 2A),and their interactions with AtObgC208e681 were analyzed through ayeast two-hybrid assay. The results show that the N-terminaldomain (NTD) of AtRSH1 combined with the TGS domain inter-acted most strongly with AtObgC208e681 (Fig. 2Bc) and indicate thatthe C-terminal domain (CTD) region lacking the TGS domainnegatively regulates this interaction. However, the deletion of theTGS domain, AtRSH1(D567e626), resulted in the lowest bindingactivity (Fig. 2Bd). Furthermore, due to the absence of the TGS re-gion, AtRSH1(82e566) also exhibited a much weaker interactionthan AtRSH1(82e635), which suggests that the TGS domain is as anenhancer region for the AtRSH1 interaction with AtObgC (Fig. 2Bband c). Moreover, the AtRSH1(489e883) construct lacking the HDand SYNTH regions and the AtRSH1(82e566) construct without theTGS region showed slightly reduced binding activity comparedwith AtRSH1(78e883) (Fig. 2Ba and e). Although the catalytic ac-tivity of the conserved HD and SYNTH domains may have no sub-stantial effects on the interaction, the full-length NTD of AtRSH1 isrequired for interaction with AtObgC (Fig. S2). These results implythat the TGS domain of AtRSH1 considerably promotes the inter-action between the NTD of AtRSH1 and AtObgC and that the C

Fig. 2. Identification of critical regions of AtRSH1 for interaction with AtObgC. The indicatedLexA BD-AtObgC in the EGY48/pSH18-34 strain. (A) Schematic representations of the varioucontaining partial AtRSH1 gene fragments are indicated (top). (B) The interactions of partial Arelative b-galactosidase activity units. The error bars represent the means � SD values obta

terminus of AtRSH1 may block the interaction with AtObgC.Therefore, these findings lead to the conclusion that the TGSdomain and the NTD of AtRSH1 are crucial for the interaction withAtObgC.

2.3. Molecular docking simulations using homology models furthersupport the AtObgCeAtRSH1 interaction

Togain a better understandingof the interactionbetweenAtObgCand AtRSH1, we performed proteineprotein docking simulationsand analyzed the binding conformations. According to our yeasttwo-hybrid assay results, the 3D structures of AtObgC208e681 andAtRSH182e635 were constructed through the homology modelingmethod (Fig. 3A). The PROCHECK and ProSA-Web programs wereused to validate and identify the overall model qualities of thehomology-modeled structures. All of the validation results revealedthat our model structures were well constructed (see Text S1).

The structural binding conformation between AtObgC208e681and AtRSH182e635 was predicted by proteineprotein docking sim-ulations using the obtained homology models (Fig. 3B). We alsoinvestigated the electrostatic potential (ESP) distributions of theindividual proteins to analyze the surface residues buried in theputative complex structures. The ESP distributions betweenAtObgC208e681 and N-terminal domain of AtRSH1 containing theTGS domain were clearly depicted by two successive 45-degreerotations (Fig. 3C). The ESP surface examination helped demon-strate that the negatively charged face of the three domains ofAtObgC208e681 interacts with the positively charged face ofAtRSH182e635.

To obtain further information on whether the TGS domain ofAtRSH1 is important for the interaction with AtObgC208e681, addi-tional docking simulations were performed between AtObgC208e

681 and AtRSH1 without the TGS domain and then compared withthe docked structure of AtObgC208e681 and AtRSH1 with the TGSdomain. One hundred poses were sorted out from the dockingsimulations in the with-the-TGS-domain case, only six poses werefiltered out in thewithout-the-TGS-domain case. It appears that theTGS domain promotes the binding stability of the AtObgC208e681e

AtRSH182e635 complex. To confirm this hypothesis, we checked theESP difference between thewith- and thewithout-the-TGS-domain

AtRSH1 domains fused to the B42AD domain were analyzed for their interaction withs truncated AtRSH1 proteins used for the yeast two-hybrid assay. Five AtRSH1 proteinstRSH1 fragment constructs (aee) with AtObgC (208e681)were compared based on theirined from three biological replicates.

Fig. 3. Proteineprotein interaction between AtObgC and AtRSH1 through the computational docking method. (A) Cartoon representation of the homology modeling structures ofAtObgC208e681 (green) and the AtRSH1 N-terminal domain with TGS (cyan). (B) Binding conformation between AtObgC and AtRSH1 as a final result. Each domain of both proteins isshown in a different color. (C) Electrostatic potential (ESP) surfaces of the docked structures of AtObgC208e681 and AtRSH1. To clarify the interface region, the structures areconsecutively rotated 45� from the complex structure, which is highlighted by the circle, to each single protein. Comparison of ESP surfaces for AtRSH1 (D) with and (E) without theTGS domain. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

J. Chen et al. / Plant Physiology and Biochemistry 74 (2014) 176e184 179

systems (Figs. 3D and E). The formal net charges of AtObgC208e681,the NTD with the TGS domain, and the NTD region without theTGS domain were found to be �6, þ5, and �1, respectively.The proteineprotein docked conformation of the NTD with TGSdomain showed an increase in positive charge, which indicatesenhanced electrostatic complementarity between AtRSH182e635

and AtObgC208e681. We also found enhanced VDW (van der Waals)interaction at the binding position in the presence of the TGSdomain. Thus, these results confirm that the N terminus of AtRSH1with the TGS domain can interact more strongly with AtObgC208e681 and highlight the importance of the TGS domain for the stabi-lization of the AtRSH1eAtObgC interaction.

J. Chen et al. / Plant Physiology and Biochemistry 74 (2014) 176e184180

In addition, while checking the interface region of the dockedconformation, all of the domains in AtObgC208e681 were interactedwith the AtRSH182e635 structure, as shown in Fig. 3B. This resultindicates that the all of the domains of AtObgC208e681 are importantfor its interaction with AtRSH1. Finally, these data providegeometrical details of a reasonable AtObgC208e681eAtRSH182e635binding conformation and support the experimental results fromstructural proteomics.

2.4. The expression patterns of AtObgC and AtRSH1 are positivelycorrelated in response to wounding and salt stresses

Wounding and salt stresses induce high ppGpp accumulation inArabidopsis chloroplasts (Takahashi et al., 2004). To furthercomprehend the physiological role of the interaction betweenAtObgC and AtRSH1 in response to ppGpp-mediated stresses, weperformed quantitative real-time PCR (qRT-PCR). After 1-hwounding treatment, the transcript levels of AtRSH1 rapidlydecreased to 77.6% compared with the untreated plants (control;Fig. 4A). This phenomenon is consistent with the recent report thatAtRSH1 is downregulated at the transcription level with woundingtreatment, and this protein is also predicted to act as a ppGpp hy-drolase and decrease ppGpp levels. Based on these findings, it isrational to hypothesize that the reduction in AtRSH1 expressionunder stressed conditions leads to the maintenance of high intra-cellular ppGpp levels (Mizusawa et al., 2008). Interestingly, theAtObgC gene expression under the same stress condition was alsoreduced to 66.2%. In addition, when seedlings were treated with100 mM NaCl for 1 h, the transcript levels of AtRSH1 and AtObgCwere also lower than those found in the untreated controls(Fig. 4B). Contrary to the expression of AtRSH1, however, AtRSH2encoding the strongest ppGpp synthetase showed increased tran-scription after wounding and high salt stress (Mizusawa et al.,2008). These results imply a positive correlation between AtObgCand AtRSH1 in the plant responses to wounding and salt stresses.

Moreover, all AtRSH genes exhibit individual temporal expres-sion by circadian clock regulation (Mizusawa et al., 2008). Tofurther investigate whether AtObgC and AtRSH1 would respond towounding and salt stresses in the circadian pattern, seedlings weretreated with wounding or 100 mM NaCl for 1 h at an interval of 8 hduring a 12-h light/12-h dark cycle (Fig. 5A). Subsequently, tem-poral expressions of AtObgC and AtRSH1were detected by qRT-PCR.

Fig. 4. Expression profiling of AtObgC and AtRSH1 in response to wounding and high saexpression in Arabidopsis rosette leaves after wounding treatment. MS-medium-grown plharvested after 1 h (gray bar). The control leaves (white bar) were not wounded. (B) ExpressMS mediumwere treated with 100 mM NaCl for 1 h. The total RNA was extracted fromwholThe white and gray bars indicate the untreated and the 100-mM-NaCl-treated plants, rerespectively. Each transcript level was normalized with that of tubulin as an internal controlmeans � SE.

Although the wounding and salt treatments did not influence thecircadian patterns of the expression levels of both genes, thetranscript levels of AtRSH1 and AtObgCweremuch lower than thoseof the untreated controls throughout the experiment (Fig. 5B andC). The results demonstrate that the expression of the AtObgC andAtRSH1 genes is diminished by both stresses, regardless of lightconditions. The expression profiling of AtObgC and AtRSH1 underthe conditions of wounding and high salinity supports the findingthat both genes are positively correlated in ppGpp-mediated stressresponses. To exclude the possibility of influences caused by os-motic stress during the beginning of wounding treatment, thetranscript levels of AtObgC, AtRSH1, and salt and wound stress-specific markers RD29A (Koprivova et al., 2008) and VSP2 (Suzaet al., 2010) a, respectively, were examined under the woundingcondition. The results should that the relative mRNA levels ofAtObgC, AtRSH1, and RD29A genes but not the VSP2 gene weredecreased, which implies a specific response of AtObgC and AtRSH1to wounding stress (Fig. S3).

Additionally, a transient expression assay and reverse-transcription PCR (RT-PCR) analysis demonstrated that AtRSH1 isa light-inducible gene encoding a chloroplast-targeted protein,similarly to AtObgC (Fig. 6) (Bang et al., 2009). In general, exposureto stress causes a highly coordinated downregulation of photo-synthesis genes to minimize photosynthesis-associated damage inplants (Less et al., 2011). Therefore, it is reasonable to conclude thatthe expression of both AtObgC and AtRSH1 genes may be reduced toallow ppGpp accumulation in order to arrest chloroplast develop-ment during various stresses and that their expressions may berecovered for AtObgCeAtRSH1 interaction to eliminate excessppGpp during normal growth after stress.

Recently, Garcia’s group proposed a different name for AtObgC,namely, CPSAR1 (chloroplast homolog of cytosolic SAR1), becauseits vesicular transport function is similar to that of the cytosolicCOPII protein SAR1, even though their phylogenetic analysisexhibited that it is related to the Obg family (Garcia et al., 2010).Despite their findings, our previous work (Bang et al., 2012) and thepresent study on ribosome biogenesis and stress responses stronglysuggest that the role of AtObgC is more similar to that of bacterialObg and not the SAR1 protein. However, because the potential roleof AtObgC in the regulation of ppGpp signaling provides flexibleresponses to various environmental changes during chloroplastdevelopment, we cannot exclude the possibility that AtObgC may

lt stresses. (A) Quantitative real-time PCR (qRT-PCR) analysis of AtObgC and AtRSH1ants were wounded by crushing the leaves with forceps, and the leaf samples wereion patterns of AtObgC and AtRSH1 under salt stress. Two-week-old seedlings grown one seedlings, and the levels of AtObgC and AtRSH1 mRNAwere also analyzed by qRT-PCR.spectively. VSP2 and RD29A were used as maker genes of wounding and salt stress,. The data are representations of three independent experiments and are plotted as the

Fig. 5. The stress-reduced expression patterns of AtObgC and AtRSH1 are circadian clock-independent. (A) Two-week-old seedlings grown on MS medium were treated withwounding or 100 mM NaCl under a 12-h light/12-h dark photoperiod. Samples were collected every 8 h after 1-h wounding or salt treatment, which is marked with an arrow underthe photoperiod bar. Total RNAwas extracted from the leaves or whole seedlings, and the levels of AtObgC and AtRSH1mRNAwere also analyzed by quantitative real-time PCR (qRT-PCR). (B) Wounding-regulated expression patterns of AtObgC and AtRSH1 in Arabidopsis rosette leaves. MS-medium-grown plants were wounded by crushing the leaves withforceps; starting 1 h after treatment, leaf samples every 8 h (gray bar). The control leaves (white bar) were not wounded. (C) Salt-stress-reduced expression patterns of AtObgC andAtRSH1 under 12-h light/12-h dark photoperiod. The white and gray bars indicate the untreated and the 100-mMeNaCl-treated plants, respectively. VSP2 and RD29A were used asmaker genes of wounding and salt stress, respectively. Each transcript level was normalized with that of tubulin as an internal control. The data are representations of three in-dependent experiments and are plotted as the means � SE.

J. Chen et al. / Plant Physiology and Biochemistry 74 (2014) 176e184 181

have multiple functions to regulate diverse physiological processes.Although bacterial Obg is essential for SpoT to hydrolyze additional(p)ppGpp under nutrient-rich conditions (Raskin et al., 2007;Woutet al., 2004), it remains unclear whether ObgC does regulate RSH1in plants. To understand these further, more mutant analyses inaddition to co-immunoprecipitation (Myouga et al., 2008) and/orbimolecular fluorescence complementation (Waadt et al., 2008;Zhang and Hu, 2010) analyses are needed. For example, doublemutants of AtObgC and AtRSH1 may provide more informationnecessary to better elucidate the correlation between their inter-action and stress responses.

3. Conclusions

In this study, we characterized the interaction between AtObgCand AtRSH1 in Arabidopsis, which are bacterial Obg and SpoT ho-mologs, respectively, in response to environmental stresses. Yeast

two-hybrid assays revealed their binding sites, which are essentialfor the interaction. Furthermore, based on the homology modelingmethod, docking simulations further supported the finding that allof theAtObgCdomains and theN terminus of AtRSH1 containing theTGS region contribute significantly to the interaction. In addition,the results of qRT-PCR analyses implied a positive correlation be-tween AtObgC and AtRSH1 in the plant responses to wounding andhigh salinity. To the best of our knowledge, this is the first report ofthe role of AtObgC in plant responses to environmental stresses.Although the direct regulation of the ppGpp level by AtObgC re-mains to be proven, it should be noted that AtObgC interacts withAtRSH1 of Arabidopsis, a bacterial SpoT homolog protein, which is aputative ppGpp hydrolase. The broad ranges of interaction betweenObg and SpoT not only in many bacterial species but also in plantsdescribed in previous reports and here may underlie the evolu-tionary conservation of the ppGpp-mediated stress signalingpathway from prokaryote to plant chloroplasts.

Fig. 6. Chloroplast targeting and gene expression pattern of AtRSH1 support the in vivo interaction between AtObgC and AtRSH1. (A) Intracellular localization of the full-lengthAtRSH1. The images exhibit the in vivo targeting of AtRSH1-GFP in protoplasts. Arabidopsis protoplasts were transformed through PEG-mediated transformation with AtRSH1-GFP and observed after 48 h. The green color indicates the expressed AtRSH1-GFP in protoplasts. The orange color in the ‘Merge’ section indicates the overlap of green(AtRSH1-GFP), red (CHL, chlorophyll), and DIC (differential interference contrast). Scale bar ¼ 10 mm. (B) The expression pattern of AtObgC was similar to that of AtRSH1. Thetranscript levels of AtObgC, AtRSH1, AtRSH2, and AtRSH3were analyzed by RT-PCR using 5 mg of total RNAs for the reverse-transcription reaction. Contrary to the AtRSH2 and AtRSH3genes, AtRSH1 showed a similar expression pattern with that of AtObgC, i.e., the AtObgC and AtRSH1 genes showed a shoot-specific and light-inducible expression pattern. Tubulinwas used as an internal control for the normalization. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

J. Chen et al. / Plant Physiology and Biochemistry 74 (2014) 176e184182

4. Materials and methods

4.1. Phylogenetic analysis

The amino acid sequences of the RelA/SpoT homologs used inthis study can be found in the GenBank protein database under thefollowing accession numbers (or locus IDs): Arabidopsis thalianaNP_849287 (AtRSH1, AT4G02260.1), NP_188021 (AtRSH2, AT3G14050.1), NP_564652 (AtRSH3, AT1G54130.1), and NP_188374(AtCRSH, AT3G17470.1); E. coli NP_417264 (RelA, b2784) andNP_418107 (SpoT, b3650). Their amino acid sequences were alignedusing the ClustalW program, and the phylogenetic tree was con-structed based on the neighbor-joining method using the MEGA5.05 software (Tamura et al., 2011). The bootstrap analysis wasconducted with 1000 replicates.

4.2. Yeast two-hybrid assay

The LexA-based Hybrid Hunter system (Invitrogen, Carlsbad, CA,USA) was used for the yeast two-hybrid assay, which was con-ducted according to the manufacturer’s protocols. For the AtObgCbait constructs, AtObgC(208e681), AtObgC(208e368), AtObgC(208e564), AtObgC(565e681), AtObgC(369e681), AtObgC(P387 V),AtObgC(S392N), and AtObgC(K393N) fragments were cloned intothe bait vectors, namely pHybLex/Zeo containing a DNA-bindingdomain. The AtRSH1(78e883) and AtRSH1(489e883) prey clones,which were constructed by fusing with the B42 activation domainof the pBAD42 vector (vanderBiezen et al., 2000), were kindlyprovided by Prof. Jonathan D. G. Jones (John Innes Centre, UK). Theother truncated AtRSH1 clones were also amplified and cloned intothe prey vector pBAD42. The prey and bait constructs were trans-formed into yeast strain EGY48/pSH18-34. The b-galactosidase ac-tivity of the resulting transformants was quantified using o-nitrophenyl-b-D-galactopyranoside (ONPG) as the substrate(Brouchon-Macari et al., 2003). Each interaction of prey and baitpairs was represented as a relative b-galactosidase activity unit oftwo interacting proteins relative to the interaction activities of thebait constructs with the prey vector only. The b-galactosidase ac-tivity was calculated according to the following equation: b-galactosidase unit ¼ 1000 � OD420/t � V � OD600.

4.3. Computational docking method

The homology model of AtRSH182e635 was generated using twotemplate structures (PDB IDs, 1VJ7 for the N-terminal domain and1NYQ for the TGS domain), and the homology model of AtObgC208e681 was contrasted using the 1LNZ and 1UDX structures as tem-plates. All of the homology modeling studies were performed usingthe MODELLER module in the Discovery Studio version 3.0 (DS 3.0)software. The homology-modeled structures were validated by twoprograms, such as PROCHECK and ProSA-web (https://prosa.services.came.sbg.ac.at/prosa.php).

To generate the binding conformation between AtObgC208e681and the N terminus of AtRSH1 with and without the TGS domain,the ZDOCK program in the DS 3.0 software was used (Chen et al.,2003). The ZDOCK program provides a complete search of all ofthe orientations between two rigid molecules through systematicrotation and translation of one molecule around another. Almost54,000 different translated orientations of AtObgC208e681 wereobtained for the stationary molecule AtRSH182e635. The additionalelectrostatic and desolvation energy termswere also used to obtainmore accurate ranking. For the site-specific docking, only thenegatively charged residues were considered to be binding siteresidues of AtObgC208e681, and only the positively charged residueslocated in the TGS domain of AtRSH182e635 were considered to bebinding site residues of AtRSH182e635. In the with-the-TGS-domaincase, the top 100 poses were categorized into 10 clusters, and thetop pose in cluster 4 (ranked 1 out of 54,000 docked poses) wasselected as the final result.

4.4. Plant treatments

Arabidopsis thaliana wild-type (Col-0) plants were MS-plate-grown under a 12-h light/12-h dark photoperiod for two weeks.The wounding treatment was performed by crushing 10% of therosette leaf surface with forceps. Wounded and non-woundedleaves were harvested 1 h after wounding (each indicated time atan interval of 8 h) and frozen in liquid nitrogen for further total RNAisolation. For the salt treatment, Col-0 seeds were grown on MSagar plates containing 1.5% sucrose at 22 �C under a 12-h light/12-hdark photoperiod. Two-week-old seedling plants were transferred

J. Chen et al. / Plant Physiology and Biochemistry 74 (2014) 176e184 183

into liquid MS media with or without 100 mM NaCl for 1 h (eachindicated time at an interval of 8 h). Salt-treated and untreatedseedlings were collected and separately frozen in liquid nitrogenfor further RNA isolation.

4.5. Quantitative real-time PCR analyses

Total RNA was extracted from the plant materials (seedlings orleaves, as indicated in the figure legends) using the TRIzol reagent(Invitrogen). Then, 2 mg of total RNA was reverse-transcribed intocDNA using the Superscript III reverse transcriptase kit (Invi-trogen). Real-time quantitative PCR was performed using theSsoFast EvaGreen Supermix and the CFX96� Real-Time System(Bio-Rad, Hercules, CA, USA). The cDNAs were amplified under thefollowing cycling conditions: (1) one cycle of 95 �C for 30 s, (2) 40cycles of 95 �C for 5 s and 55 �C for 5 s, and (3) melt-curve from65 �C to 95 �C for 5 s/step with a 0.5 �C increment. The sequences ofgene-specific primers used for the amplifications are listed inTable S1.

4.6. Transient expression of GFP-fusion proteins

The Arabidopsis (Arabidopsis thaliana) plants used for protoplastpreparation were grown on agar plates containing 2.1 g/l MS saltand 1% (w/v) sucrose. The AtRSH1-GFP fusion construct (20 mg) wasintroduced into Arabidopsis protoplasts through the polyethyleneglycol-mediated transformation method, as described previously(36). The transformed protoplasts were incubated at 22 �C in thedark. The expression of the fusion proteins was observed two daysafter transformation using an Olympus AX-70 fluorescence micro-scope (Olympus, Tokyo, Japan), and the images were captured witha cooled charge-coupled device camera (Olympus DP-70). The filtersets used were XF116-2 (exciter, 475AF20; dichroic, 500DRLP;emitter, 510AF23) and XF137 (exciter, 540AF30; dichroic, 570DRLP;emitter, 585ALP) (Omega, Inc., Brattleboro, VT, USA) for the detec-tion of the green fluorescent protein and the autofluorescence ofchlorophyll, respectively.

4.7. Reverse-transcription PCR analysis

Total RNAs were extracted from one-week-old Arabidopsisplants using the TRIzol reagent (Invitrogen) according to themanufacturer’s instructions, and 5 mg of total RNAwas used for theRT-PCR analysis. The cDNAs were constructed using the M-MuLVreverse transcriptase (Fermentas, Burlington, Canada) and an oli-go(dT18) primer, and the amplifications were performed with thecDNAs and specific primer pairs for 25 cycles (AtObgC, AtRSH2,AtRSH3), 35 cycles (AtRSH1), and 30 cycles (tubulin) of denaturationat 95 �C for 20 s, annealing at 55 �C for 30 s, and synthesis at 72 �Cfor 30 s. The PCR primer sets are listed in Table S1.

Acknowledgments

We thank Nengah Suwastika and Prof. Takashi Shiina at theKyoto Prefectural University in Japan for their discussion. D.W. Kimand S. Akhter are graduate students supported by scholarships fromthe GSNU-BK21 program, which is sponsored by theMEST in Korea.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.plaphy.2013.10.022

References

Atkinson, G.C., Tenson, T., Hauryliuk, V., 2011. The RelA/SpoT homolog (RSH) su-perfamily: distribution and functional evolution of ppGpp synthetases andhydrolases across the tree of life. PloS One 6, e23479.

Bang, W.Y., Hata, A., Jeong, I.S., Umeda, T., Masuda, T., Chen, J., Yoko, I.,Suwastika, I.N., Kim, D.W., Im, C.H., Lee, B.H., Lee, Y., Lee, K.W., Shiina, T.,Bahk, J.D., 2009. AtObgC, a plant ortholog of bacterial Obg, is a chloroplast-targeting GTPase essential for early embryogenesis. Plant Mole. Biol. 71, 379e390.

Bang, W.Y., Chen, J., Jeong, I.S., Kim, S.W., Kim, C.W., Jung, H.S., Lee, K.H., Kweon, H.S.,Yoko, I., Shiina, T., Bahk, J.D., 2012. Functional characterization of ObgC inribosome biogenesis during chloroplast development. Plant J. Cell Mol. Biol. 71,122e134.

Battesti, A., Bouveret, E., 2006. Acyl carrier protein/SpoT interaction, the switchlinking SpoT-dependent stress response to fatty acid metabolism. Mol. Micro-biol. 62, 1048e1063.

Braeken, K., Moris, M., Daniels, R., Vanderleyden, J., Michiels, J., 2006. New horizonsfor (p)ppGpp in bacterial and plant physiology. Trends Microbiol. 14, 45e54.

Brouchon-Macari, L., Joseph, M.C., Dagher, M.C., 2003. Highly sensitive microplatebeta-galactosidase assay for yeast two-hybrid systems. BioTechniques 35, 446e448.

Chen, R., Li, L., Weng, Z., 2003. ZDOCK: an initial-stage protein-docking algorithm.Prot. Struct. Function Bioinforma. 52, 80e87.

Chigri, F., Sippel, C., Kolb, M., Vothknecht, U.C., 2009. Arabidopsis OBG-like GTPase(AtOBGL) is localized in chloroplasts and has an essential function in embryodevelopment. Mol. Plant 2, 1373e1383.

Garcia, C., Khan, N.Z., Nannmark, U., Aronsson, H., 2010. The chloroplast proteinCPSAR1, dually localized in the stroma and the inner envelope membrane, isinvolved in thylakoid biogenesis. Plant J. Cell Mol. Biol. 63, 73e85.

Givens, R.M., Lin, M.H., Taylor, D.J., Mechold, U., Berry, J.O., Hernandez, V.J., 2004.Inducible expression, enzymatic activity, and origin of higher plant homologuesof bacterial RelA/SpoT stress proteins in Nicotiana tabacum. J. Biol. Chem. 279,7495e7504.

Jiang, M., Sullivan, S.M., Wout, P.K., Maddock, J.R., 2007. G-protein control of theribosome-associated stress response protein SpoT. J. Bacteriol. 189, 6140e6147.

Kasai, K., Usami, S., Yamada, T., Endo, Y., Ochi, K., Tozawa, Y., 2002. A RelA-SpoThomolog (Cr-RSH) identified in Chlamydomonas reinhardtii generates stringentfactor in vivo and localizes to chloroplasts in vitro. Nucleic Acids Res. 30, 4985e4992.

Koprivova, A., North, K.A., Kopriva, S., 2008. Complex signaling network in regula-tion of adenosine 5’-phosphosulfate reductase by salt stress in Arabidopsisroots. Plant Physiol. 146, 1408e1420.

Less, H., Angelovici, R., Tzin, V., Galili, G., 2011. Coordinated gene networks regu-lating Arabidopsis plant metabolism in response to various stresses andnutritional cues. Plant Cell 23, 1264e1271.

Maddock, J., Bhatt, A., Koch, M., Skidmore, J., 1997. Identification of an essentialCaulobacter crescentus gene encoding a member of the Obg family of GTP-binding proteins. J. Bacteriol. 179, 6426e6431.

Masuda, S., Mizusawa, K., Narisawa, T., Tozawa, Y., Ohta, H., Takamiya, K., 2008. Thebacterial stringent response, conserved in chloroplasts, controls plant fertil-ization. Plant Cell Physiol. 49, 135e141.

Mittenhuber, G., 2001. Comparative genomics and evolution of genes encodingbacterial (p)ppGpp synthetases/hydrolases (the Rel, RelA and SpoT proteins).J. Mol. Microbiol. Biotechnol. 3, 585e600.

Mizusawa, K., Masuda, S., Ohta, H., 2008. Expression profiling of four RelA/SpoT-likeproteins, homologues of bacterial stringent factors, in Arabidopsis thaliana.Planta 228, 553e562.

Myouga, F., Hosoda, C., Umezawa, T., Iizumi, H., Kuromori, T., Motohashi, R.,Shono, Y., Nagata, N., Ikeuchi, M., Shinozaki, K., 2008. A heterocomplex of ironsuperoxide dismutases defends chloroplast nucleoids against oxidative stressand is essential for chloroplast development in Arabidopsis. Plant Cell 20,3148e3162.

Okamoto, S., Ochi, K., 1998. An essential GTP-binding protein functions as a regu-lator for differentiation in Streptomyces coelicolor. Mol. Microbiol. 30, 107e119.

Potrykus, K., Cashel, M., 2008. (p)ppGpp: still magical? Annu. Rev. Microbiol. 62,35e51.

Raskin, D.M., Judson, N., Mekalanos, J.J., 2007. Regulation of the stringent responseis the essential function of the conserved bacterial G protein CgtA in Vibriocholerae. Proc. Natl. Acad. Sci. U. S. A. 104, 4636e4641.

Slominska, M., Konopa, G., Wegrzyn, G., Czyz, A., 2002. Impaired chromosomepartitioning and synchronization of DNA replication initiation in an insertionalmutant in the Vibrio harveyi cgtA gene coding for a common GTP-bindingprotein. Biochem. J. 362, 579e584.

Suza, W.P., Avila, C.A., Carruthers, K., Kulkarni, S., Goggin, F.L., Lorence, A., 2010.Exploring the impact of wounding and jasmonates on ascorbate metabolism.Plant physiol. Biochem. PPB/Soc. Fr. Physiol. Veg. 48, 337e350.

Takahashi, K., Kasai, K., Ochi, K., 2004. Identification of the bacterial alarmoneguanosine 5’-diphosphate 3’-diphosphate (ppGpp) in plants. Proc. Natl. Acad.Sci. U. S. A. 101, 4320e4324.

Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., Kumar, S., 2011. MEGA5:molecular evolutionary genetics analysis using maximum likelihood, evolu-tionary distance, and maximum parsimony methods. Mol. Biol. Evol. 28, 2731e2739.

J. Chen et al. / Plant Physiology and Biochemistry 74 (2014) 176e184184

Tozawa, Y., Nozawa, A., Kanno, T., Narisawa, T., Masuda, S., Kasai, K., Nanamiya, H.,2007. Calcium-activated (p)ppGpp synthetase in chloroplasts of land plants.J. Biol. Chem. 282, 35536e35545.

van der Biezen, E.A., Sun, J., Coleman, M.J., Bibb, M.J., Jones, J.D., 2000. ArabidopsisRelA/SpoT homologs implicate (p)ppGpp in plant signaling. Proc. Natl. Acad. Sci.U. S. A. 97, 3747e3752.

Waadt, R., Schmidt, L.K., Lohse, M., Hashimoto, K., Bock, R., Kudla, J., 2008. Multi-color bimolecular fluorescence complementation reveals simultaneous forma-tion of alternative CBL/CIPK complexes in planta. Plant J. Cell Mol. Biol. 56, 505e516.

Wolf, Y.I., Aravind, L., Grishin, N.V., Koonin, E.V., 1999. Evolution of aminoacyl-tRNAsynthetaseseanalysis of unique domain architectures and phylogenetic trees

reveals a complex history of horizontal gene transfer events. Genome Res. 9,689e710.

Wout, P., Pu, K., Sullivan, S.M., Reese, V., Zhou, S., Lin, B., Maddock, J.R., 2004. TheEscherichia coli GTPase CgtAE cofractionates with the 50S ribosomal subunitand interacts with SpoT, a ppGpp synthetase/hydrolase. J. Bacteriol. 186, 5249e5257.

Xiao, H., Kalman, M., Ikehara, K., Zemel, S., Glaser, G., Cashel, M., 1991. Residualguanosine 3’,5’-bispyrophosphate synthetic activity of relA null mutants can beeliminated by spoT null mutations. J. Biol. Chem. 266, 5980e5990.

Zhang, X., Hu, J., 2010. The Arabidopsis chloroplast division protein DYNAMIN-RELATED PROTEIN5B also mediates peroxisome division. Plant Cell 22, 431e442.