Ca

La

b

a

ARRAA

KEHT

1

mtala[elsclsiht

kf

h0

Plant Science 236 (2015) 37–43

Contents lists available at ScienceDirect

Plant Science

j ourna l ho me pa ge: www.elsev ier .com/ locate /p lantsc i

onstitutively expressed ERF-VII transcription factors redundantlyctivate the core anaerobic response in Arabidopsis thaliana

iem T. Buia, Beatrice Giuntoli a, Monika Kosmaczb, Sandro Parlanti a, Francesco Licausia,∗

Plantlab, Institute of Life Sciences, Scuola Superiore Sant’Anna, Via Mariscoglio 34, 56124, Pisa, ItalyMax Planck Institute of Molecular Plant Physiology, Am Muehlenberg 1, 14476 Golm, Germany

r t i c l e i n f o

rticle history:eceived 29 October 2014eceived in revised form 12 March 2015ccepted 13 March 2015vailable online 21 March 2015

eywords:

a b s t r a c t

Plant adaptation to hypoxic conditions is mediated by the transcriptional activation of genes involvedin the metabolic reprogramming of plant cells to cope with reduced oxygen availability. Recent studiesindicated that members of the group VII of the Ethylene Responsive Transcription Factor (ERFs) familyact as positive regulators of this molecular response. In the current study, the five ERF-VII transcriptionfactors of Arabidopsis thaliana were compared to infer a hierarchy in their role with respect to the anaer-obic response. When the activity of each transcription factor was tested on a set of hypoxia-responsive

thylene responsive factorsypoxiaranscription activation

promoters, RAP2.2, RAP2.3 and RAP2.12 appeared to be the most powerful activators. RAP2.12 was fur-ther dissected in transactivation assays in Arabidopsis protoplasts to identify responsible regions fortranscriptional activation. An ultimate C-terminal motif was identified as sufficient to drive gene tran-scription. Finally, using realtime RT-PCR in single and double mutants for the corresponding genes, weconfirmed that RAP2.2 and RAP2.12 exert major control upon the anaerobic response.

© 2015 Elsevier Ireland Ltd. All rights reserved.

. Introduction

When oxygen demand surpasses its availability in the environ-ent, plants need to adapt their metabolism and, consequently,

heir growth and developmental programs. Most of the short-nd long-term adaptations are controlled at the transcriptionalevel [1] and [2], although post-transcriptional regulation alsollows rapid switching between aerobic and anaerobic metabolism3]. Cross-species and cell-specific surveys of differential genexpression revealed the existence of a set of genes whose regu-ation pattern is conserved ubiquitously in higher plants [4]. Theseo-called core-response genes are also associated to polysomalomplexes under hypoxic conditions, indicating that active trans-ation occurs to produce the corresponding proteins [3]. Althoughome of these genes do not have a defined function yet, enzymes

nvolved in fermentation and alternative nitrite/nitrate respirationave been included in this list, as well as the transcription fac-ors and regulatory proteins involved in their control, often via

∗ Corresponding author. Tel.: +39 0502216553.E-mail addresses: t.bui@sssup.it (L.T. Bui), b.giuntoli@sssup.it (B. Giuntoli),

osmacz@mpimp-golm.mpg.de (M. Kosmacz), s.parlanti@sssup.it (S. Parlanti),.licausi@sssup.it (F. Licausi).

ttp://dx.doi.org/10.1016/j.plantsci.2015.03.008168-9452/© 2015 Elsevier Ireland Ltd. All rights reserved.

feed-back and feed-forward mechanisms [5], [6] and [7]. Anadditional set of genes, corresponding to heat shock proteinsand oxidative stress alleviation, is activated when the oxygenconcentration drops to a level at which the mitochondrial elec-tron transport is impaired and reactive oxygen species (ROS) arereleased [8] and [9].

Previously, the core-response genes in Arabidopsis thaliana havebeen shown to be activated by the ERF-VII group of the EthyleneResponse Factor family, whose stability is controlled by the N-endrule pathway for protein degradation [10] and [11]. This proteolyticpathway uses oxygen as a co-substrate to generate a degradationtag at the N-termini of the ERF-VII proteins via the concerted actionof Plant Cysteine Oxidases (PCOs), Arginyl-transferases (ATEs) andE3 ubiquitin ligases (PRT6) [12]. The Arabidopsis genome encodesfive ERF-VII group members, namely RAP2.2, RAP2.3, RAP2.12, HRE1,and HRE2. While HRE1 and HRE2 are expressed at low levels underaerobic conditions and strongly up-regulated by hypoxia, RAP2.12mRNA is ubiquitously detectable throughout development. An arti-ficially stabilized version of RAP2.12 was described as sufficient foractivating the anaerobic response in Arabidopsis [11]. The closest

RAP2.12 homologue, RAP2.2, has been suggested to be function-ally redundant in the induction of the anaerobic gene expression[13] and [6]. Indeed, Arabidopsis plants expressing an artificialmiRNA able to target both these two ERF-VII transcripts exhibited

3 Scienc

agarttvuacetbsqerA

2

2

beaAsvtDmSSdwaw(t(as

2

twPcgSgg3

2

[ra

ERF-VII proteins

Previously, the role of individual ERF-VII transcription factorsin the activation of the anaerobic response in Arabidopsis was

Fig. 1. Transactivation capacity of ERF-VII transcription factors from A. thaliana.The full-length coding sequence of each ERF-VII gene from Arabidopsis was cloneddownstream of eGFP, as depicted in the scheme, to generate an effector plasmid.Arabidopsis mesophyll protoplasts were transiently transformed with a plasmidbearing a Firefly Luciferase (FLuc) gene under control of a hypoxia-responsive

8 L.T. Bui et al. / Plant

reduction in the up-regulation of the core anaerobic responseenes [11]. Additionally, a double hre1hre2 mutant did not displaylterations during the first phases of the induction of the anaerobicesponse, leading us to speculate that HREs may play a role in sus-aining active transcription [14] after activation of the anaerobicranscriptional program by RAP2.12 and, possibly, RAP2.2. Con-ersely, the role of RAP2.3 in the anaerobic response remainednexplored; whereas this transcription factor has been recentlyssociated to the regulation of NO-mediated physiological pro-esses, including germination, stomata opening and hypocotyllongation [15]. In the current study, we further characterizedhe trans-activation properties of the ERF-VII proteins in Ara-idopsis, concluding that RAP2.2, RAP2.3 and RAP2.12 possess thetrongest activation capacity of anaerobic promoters. We subse-uently exploited RAP2.2 and RAP2.12 T-DNA mutants to gathervidence that these two transcription factors are sufficient andequired to redundantly activate the core anaerobic response inrabidopsis.

. Material and methods

.1. Plant materials and growth conditions

The Columbia-0 (Col-0) and Wassiljewska (Ws) ecotype of Ara-idopsis thaliana were used as wild-type backgrounds in all thexperiments. The T-DNA insertion mutants rap2.2 (SAIL 18 G09)nd rap2.12 (FLAG 525G09) were obtained from the Europeanrabidopsis Stock Centre (NASC) and the Versailles Arabidop-is Stock Center, respectively. Homozygous lines were identifiedia PCR screening of genomic DNA using gene-specific primersogether with T-DNA-specific primers (Supplemental Table S1).ouble homozygous lines were obtained by crossing the two singleutants and then screening the F2 generation as described above.

eeds were stratified at 4 ◦C in soil (1/3 perlite, 2/3 Hawitaflor®

pezialsubstrat) in the dark for 48 h and germinated at 22 ◦Cay/18 ◦C night with a 12-h photoperiod. The quantum irradianceas 100 �mol photons m−2 s−1. Oxygen treatments were applied

s described in [16] and treating plants or protoplasts in the darkith normal air containing 21% oxygen (normoxia) or 1% oxygen

hypoxia) for the time indicated in the figure legends. All plantreatments started two hours after the beginning of the light phase10.00 a.m.). Protoplasts were treated at the same time of day (10.00.m.), although they were maintained in darkness also before expo-ure to hypoxia.

.2. Construct design and production

Construction of non-binary high copy number vectors forhe transient expression of the five ERF-VII transcription factorsas described in [6]. The reporter plasmids PromHb1:FLuc and

romHRA1:FLuc were described in [7] and [11]. The PromSAD6:FLuconstruct, instead, was generated by PCR-amplification of a 1244 bpenomic fragment corresponding to the 5′ region upstream of theAD6 (At1g43800) coding sequence from the Arabidopsis Col-0enome, its cloning into the pENTR/Topo vector (Life Technolo-ies), and subsequent recombination into a pGWL7 vector [17]. A5S:RLuc vector was used as the normalization control [18].

.3. Trans-activation assays using protoplasts

Arabidopsis mesophyll protoplasts were prepared according to19] and transformed using 3 �g of each plasmid. A dual luciferaseeporter assay was performed as described previously [18]. Allssays were performed with at least four replicates and repeated

e 236 (2015) 37–43

twice. Average relative signals are shown, together with their rel-ative standard deviation.

2.4. Gene expression analyses

RNA was extracted from 5-week old plants. Total RNA, extractedas described by Kosmacz et al. [20], was subjected to DNasetreatment using the RQ1-DNase kit (Promega). Five hundred ngRNAwere reverse transcribed into cDNA using the iScriptTM cDNASynthesis Kit (Biorad). Real-time PCR amplification was carriedout with the ABI Prism 7300 sequence detection system (AppliedBiosystems), using a iQSYBR Green Supermix (Biorad). Ubiquitin10(At4g05320) was used as the housekeeping gene. The primer pairsused in the realtime qRT-PCR were previously described in [11].Relative quantification of the expression of each individual genewas performed using the 2−��C(T) method [21].

3. Results

3.1. Assessment of the transactivation ability of the Arabidopsis

promoter (PromHRA1, PromSAD6, or PromHb1), an effector plasmid for ERF-VIIexpression and a normalization plasmid (35S:RLuc). The level of promoter activityin the absence of any effector was set to 1. Data are presented as mean ± s.d. (n = 4),asterisks indicate statistically significant significance (P < 0.05, one-way ANOVA,Holm-Sidak post-hoc test).

Scienc

sttprcCH(battfwttoudta

Fdatact

L.T. Bui et al. / Plant

tudied using transgenic approaches. With the aim of comparinghe activation potency of all five Arabidopsis proteins belongingo this group, we designed a trans-activation assay in mesophyllrotoplasts by fusing anaerobic promoters with a Firefly Luciferaseeporter gene (Prom:FLuc). Three different promoter regions werehosen, upstream of the genes coding for the STEAROYL-ACYLARRIER PROTEIN �9-DESATURASE6 (SAD6), the non symbiotic-emoglobin (Hb1), and the Hypoxia Responsive Attenuator 1

HRA1). The induction of PromHRA1 by low oxygen conditions haseen shown by Giuntoli et al. [7], while the potential of promHB1nd promSAD6 to confer hypoxia-responsiveness was validated inhe current study, by exposing Arabidopsis protoplasts, transientlyransformed with both Prom:Fluc constructs, to hypoxia (1% O2)or 6 h (supplemental figure S2). Additionally, these three genesere previously identified as potential targets of RAP2.12 in whole-

ranscriptome or gene-targeted surveys [7] and [11]. The protoplastransactivation assay was carried out by analyzing the expressionf the FLuc reporter gene, measured as luciferase enzymatic activitypon supply of the luciferin substrate, under control of the three

ifferent promoters, either alone or in the presence of one ERF-VIIranscription factor at a time. The effector protein was produced by

second high-copy plasmid, co-transfected in protoplasts together

ig. 2. mRNA abundance of RAP2.2, RAP2.3 and RAP2.12 in Arabidopsis tissues over planifferent developmental stages. The “medium” expression level corresponds to the interqnd “High” expression levels correspond to the first and the fourth quartiles respectively. (bissues. (c) Abundance of mRNAs associated to immunopurified ribosomes expressed unnd RAP2.12 transcripts in translatomes isolated from different cell populations were plotorresponding tissues, and the promoter exploited for cell-type discrimination, is given

he eFP platform (http://efp.ucr.edu/cgi-bin/absolute.cgi). Absolute signal values of transc

e 236 (2015) 37–43 39

with the reporter (prom:FLuc) and normalization (35S:RLuc) plas-mids. Since the N-terminal consensus, shared among the ERF-VIIproteins, leads to continuous proteolysis in the presence of oxy-gen and nitric oxide [6] and [15], the proteins were stabilized byadding a GFP sequence at their N-termini that masked the consen-sus and thereby hindered the oxidation of the terminal cysteineresidue by PCOs. The florescence signal emitted by GFP was alsoused to confirm the success of protoplast transformation and actualexpression and nuclear localization of the transcription factors(supplemental Fig. S3). Our trans-activation assay revealed thatRAP2.2, RAP2.3 and RAP2.12 possess the strongest ability to acti-vate the three anaerobic reporters (Fig. 1). HRE1 and HRE2, instead,did not exhibit significant trans-activation activity (Fig. 1).Consti-tutive ERF-VII genes display partially redundant and overlappingmRNA distribution.

Neither HRE1 nor HRE2 alone are able, when overexpressed, toup-regulate the anaerobic response (Fig. 1), nor are their mRNAsreadily available for translation at the onset of hypoxia, given thatboth HRE1 and HRE2 are expressed at very low levels under aero-

bic conditions [14]. These observations hint at a major role of theother three members of this group in ERF-VII mediated inductionof the hypoxia-responsive genes. However, the relevance of their

t development. (a) Average mRNA abundance of RAP2.12, RAP2.3 and RAP2.12 atuartile region of the overall expression for all probesets on the ATH1 array. “Low”) Average mRNA abundance of RAP2.12, RAP2.3 and RAP2.12 in different Arabidopsis

der the control of cell-specific promoters. Absolute signal values of RAP2.2, RAP2.3,ted as a heat-map onto schematic drawings of a leaf and root sections. A key for thein the top part of the panel. Images were edited from the originals extracted fromripts in translatomes isolated from cell populations are shown.

4 Science 236 (2015) 37–43

cagslattpieeRalmmpRpRlEo

3e

tsgviaAitsritCaaasRtBttm

rcprrTooRofb

Fig. 3. Identification of the activation domain(s) contained in the RAP2.12 protein.(a) Schematic overview of the conserved domains identified in the ERF-VII proteinsof A. thaliana, following the nomenclature assigned by Nakano et al. [25]. (b) Trans-activation capacity of different RAP2.2 and RAP2.12 C-terminal fragments, fused tothe GAL4 DNA binding domain, in Arabidopsis mesophyll protoplasts co-transfectedwith a plasmid bearing the FLuc reporter under control of a GAL4-responsive pro-moter. A fusion between the GUS protein and GAL4 DNA binding domain was used asa negative control. Data are presented as mean ± s.d. (n = 4), asterisks indicate sta-

0 L.T. Bui et al. / Plant

ontribution to the actual activation of the molecular response tonaerobiosis is curtailed to the cell types where these three ERF-VIIenes are actually expressed. We therefore analyzed the expres-ion level of the RAP2.2, RAP2.3 and RAP2.12 genes exploiting thearge amount of transcript data available at the Genevestigator [22]nd BAR [23] databases. Expression of these three genes appearedo be present throughout all stages of Arabidopsis development inhe fourth quartile of the overall signal intensity recorded for allrobe sets present on the Affymetrix ATH1 chip (Fig. 2a). A closer

nvestigation at the tissue level indicated that, although differencesxist in the expression of the three genes, at least one of them isxpressed at medium level in each sample (Fig. 2b). Overall, RAP2.2,AP2.3 and RAP2.12 are expressed in each tissue considered in thenalysis. In few cases, such as pollen and phloem tissues, the mRNAevel of one or two ERF-VII gene decreases, whereas the third gene

aintains moderate to high expression (Fig. 2b). Differences inRNA abundance were also observed when transcriptome-wide

olysome loading was assessed in different cell populations [24].AP2.3 and RAP2.2 transcripts exhibit higher polysome loading inhotosynthetically active tissues and ion the epidermis, whereasAP2.12 is associated to ribosomes in the root cortex and vascu-

ature (Fig. 2c). Taking these observations into consideration, eachRF-VII gene exhibits the potential to contribute to the activationf the anaerobic response in different tissues.

.2. Transcriptional activation is provided by multiple activatinglements in RAP2.12

The observation that RAP2.2, RAP2.3and RAP2.12 can triggerhe most pronounced induction of anaerobic promoters (Fig. 1)uggested that, among all ERF-VII encoded in the Arabidopsisenome, these three possess the structural requisites to acti-ate the molecular response to anaerobiosis. With the scope ofdentifying protein domains responsible for the observed trans-ctivation ability, we compared the aminoacid sequences of therabidopsis ERF-VII, looking for the conserved motifs (CMVIIs)

dentified by Nakano et al. [25] as distinctive of this group ofranscription factors. All ERF-VII possess the N-terminal consen-us MCGGAI(I/L) (CMVII-1), the AP2 domain responsible for DNAecognition and binding, and an additional motif whose functions still unknown (CMVII-3). RAP2.2 and RAP2.12 contain most ofhe conserved motifs, including four conserved sequences at the-termini (Fig. 3a).A comparison of such conserved motifs pointedt an involvement of CMVII-5, present in all members except HRE2,nd either CMVII-8 (present in RAP2.12 and HRE1) or CMVII-7nd CMVII-4 (unique of RAP2.2 and RAP2.12) in determining theuperior transactivation potential displayed byRAP2.2, RAP2.3 andAP2.12. CMVII-5 corresponds to a LWSY motif identified at the C-erminus of other ERF proteins, including Cold Responsive Elementinding Factors (CBFs) [26]. Interestingly, RAP2.3 does not con-ain C-terminal conserved motifs, suggesting that the mechanismhrough which it promotes transcription relies on a non-conserved

otif.Previously, a transactivation assay in mesophyll protoplasts

evealed that the overall transactivation properties of RAP2.12ould be circumscribed to the ultimate C-terminal region of therotein. Indeed, deletion of the last 18 RAP2.12 aminoacids stronglyeduced the transactivation activity of the transcription factor, andemoval of additional 40 aminoacids completely abolished it [11].o gather experimental evidence for the autonomous contributionf each of the conserved regions to the transactivating propertiesf the ERF-VII transcription factors, C-terminal fragments of

AP2.12 were used in chimeric fusion with a DNA binding domainf the yeast (Saccharomyces cerevisiae) GAL4 (G4D) transcriptionactor. RAP2.12 was chosen as the paradigm for the ERF-VII groupecause it contains all the CMVII identified by [25], whereas

tistically significant significance (P < 0.05, one-way ANOVA, Holm-Sidak post-hoctest).

�-glucuronidase (GUS) was fused to G4D to be used as a negativecontrol. The effect of these chimeric effectors on a UAS:FLucreporter was evaluated (Fig. 3b). Activation of FLuc was observedfor all fragments tested, with one exception: a fragment of RAP2.12spanning from the end of the DNA Binding Domain (DBD) until

the 254th aminoacid, and bearing two conserved motifs (CMVII-5and CMVII-8), strongly hampered the activation of the reporter(Fig. 3b). A corresponding fragment of RAP2.2 (266–379) also

L.T. Bui et al. / Plant Science 236 (2015) 37–43 41

Fig. 4. Genotyping of rap2.2, rap2.12 and rap2.2-12 mutants. (a) Position of the T-DNA insertion (thick arrows) in the rap2.2 (At3g14230) and rap2.12 (At1g53910) mutant linesand location of the annealing sequence for the PCR primers used for the screen (thin arrows). (b) The result of a PCR screen is displayed, as analyzed using gel electrophoresis.(c) mRNA levels of RAP2.2 and RAP2.12 in rap2.2(Sail 184 G12), rap2.12(FLAG 525G09), and the respective wild-type ecotypes (Columbia-0 and Wassiljiewska) under aerobicand hypoxic conditions (1% O2, 3 h). Data are presented as mean ± s.d. (n = 4) and referred to the aerobic wild-type samples (wt, air = 1). Asterisks indicate significant differences(P < 0.05, t-test within treatments). (d) Detection of full length coding sequences of RAP2.2 and RAP2.12 in two different rap2.2-12 double mutant plants. An aerobic wild-typec luded

emiftawfm

3r

tpaniaaktm((l(

DNA and a plasmid containing each coding sequence (“Pos. control”) were also inc

xhibited the same inability to activate transcription, while ainimal part (361–379), containing the sole CMVII-5, was able to

nduce FLuc expression (Fig. 3b). Although a biological explanationor these observations is not straightforward, it can be speculatedhat direct linkage of CMVII-4 to G4D, or occurrence of CMVII-4nd CMVII-5 alone, inhibit transcription activation. Nevertheless,ith this experiment we obtained evidence that a short C-terminal

ragment of RAP2.2 or RAP2.12, containing the conserved LWSYotif, can be used as a minimal transcription activation domain.

.3. RAP2.12 and RAP2.2 redundantly activate the anaerobicesponse

Both RAP2.2 and RAP2.12 have been shown to be able to sus-ain the induction of anaerobic genes, when overexpressed inlanta [13] and [14], while overexpression of an artificial microRNAgainst both genes has the opposite effect. However, these data can-ot exclude that other transcription factors are required for the fast

nduction of the anaerobic genes upon hypoxia. If, indeed, RAP2.2nd RAP2.12 are the sole or major regulators responsible for thectivation of the molecular response to anaerobiosis, concomitantnocking-out of both proteins should abolish this process. To testhis hypothesis, we searched the public collections for Arabidopsis

utants bearing T-DNA insertions inside each locus: a T-DNA line

Sail 184 G12) in the Columbia-0 ecotype was selected for RAP2.2At3g14230), while a mutant line (FLAG 525G09) from the FLAG col-ection in the Wasiljieska background was identified for RAP2.12At1g53910) (Fig. 4a). After validating the position of the T-DNA

as controls. The arrows indicate the 1 kb reference band on the DNA marker (M).

insertions inside each gene (Fig. 4b), we confirmed the reduc-tion of RAP2.2 and RAP2.12 expression in the respective mutants(Fig. 4c). Moreover, both lines were crossed to generate a rap2.2-12double mutant. Neither single nor double homozygous mutantsdisplayed any distinct phenotype when compared to either of thetwo parental ecotypes (data not shown). However, it should benoted that, whereas the T-DNA insertion in FLAG 525G09 inter-rupted the coding sequence, in Sail 184 G12 a full-length CDS couldstill be amplified using cDNA as a template after 40 PCR cycles(Fig. 4d): this would explain the behavior of the Sail 184 G12 lineas a knock-down mutant (Fig. 4c), but also imply that RAP2.2 is stillexpressed in the double mutant line, although at much lower levelsthan normal.

Subsequently, we analyzed the expression of the hypoxiamarker genes Alcohol Dehydrogenase (ADH), Pyruvate Decarboxylase1 (PDC1), Hemoglobin (Hb1) and Plant Cysteine Oxidase 1 (PCO1) inthe single mutant genotypes. No significant decrease in the hypoxicresponse could be detected in the single mutants when compared tothe respective wild-types (Fig. 5a). Instead, homozygous insertionof both T-DNAs caused a significant reduction in the induction ofthe anaerobic targets (Fig. 5b), when compared to a cross betweenthe Col-0 and Ws ecotypes. From these results, we concluded thatRAP2.2 and RAP2.12 redundantly control the molecular responseto anaerobiosis in Arabidopsis. The residual activation of the HB1

gene tested in the double T-DNA line could be explained eitherby the contribution of RAP2.3 or by small amounts of RAP2.2 stillbeing produced in the rap2.2 T-DNA line and its double mutantprogeny.

42 L.T. Bui et al. / Plant Science 236 (2015) 37–43

Fig. 5. Differential expression of hypoxia-responsive genes, measured by realtime qRT-PCR, in rap2.2 and rap2.12 mutants as compared with their respective wild ecotypes.(a) Relative mRNA level measured in 5-week old rosettes treated under normoxic (empty bars) or hypoxic conditions (1% O2, 3 h, black bars) in Columbia-0, Wassiljeska,rap2.2 (Col-0 background), rap2.12 (Ws background). Data, normalized to the Col-0 ecotype in air, are presented as mean ± s.d. (n = 4). (b) Relative mRNA levels in 5-weekold rosettes treated under normoxic or hypoxic conditions in a cross (F1 generation) between Col-0 and Ws (gray bars), and a double mutant rap2.2-12 (white bars). Data,n t diffeu

4

tw[brasuoew

ormalized to the wild ecotype in air, are presented as mean ± s.d. (n = 4). Significansing a t-test, are indicated with an asterisk.

. Discussion and conclusions

The transcriptional response to low oxygen conditions in A.haliana consists in the conserved induction of about fifty genes,hose orthologs are also induced by hypoxia in other plant species

4]. The group VII of the ERF family of transcription factors haseen suggested to be involved in the control of this molecularesponse [10], [11], [13] and [14]. The hypoxia-inducible HRE1nd HRE2 ERF-VIIs were shown to be responsible to sustain tran-cription of the anaerobic genes, although unable to initiate their

p-regulation [14]. This hypothesis was initially based on the lackf activation of the anaerobic response in transgenic plants over-xpressing either HRE1 or HRE2. These transgenic plants expressedild-type versions of the ERF-VIIs, which are subjected to rapid

rences between each mutant and its respective wild type under hypoxia, evaluated

proteolysis via the N-end rule pathway [11], therefore it could stillbe hypothesized that lack of protein stabilization and accumula-tion under normoxic conditions was responsible for not triggeringthe induction of target genes. However, our new results usingstabilized versions of all ERF-VII on typical hypoxia-responsive pro-moters in mesophyll protoplasts indicated that HRE1 and HRE2do not possess trans-activation capacity alone. Instead, RAP2.2,RAP2.3 and RAP2.12 exhibited a strong positive effect on theregulation of all anaerobic promoters tested (Fig. 1). The impor-tance of HRE1 and HRE2 in the initial activation of the anaerobic

response is further diminished by the fact that their respectivegenes are expressed at very low level under aerobic conditions andinduced by hypoxia, likely under the control of RAP2.2 and RAP2.12[11].

Scienc

laareiiptVhomE

vtwmataRtt

aatilRmgaietaAcwa

A

Soi

A

fp

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

L.T. Bui et al. / Plant

The 49 genes that constitute the core molecular response toow oxygen conditions, as defined above, are induced in a cellutonomous manner, therefore every cell is expected to possess

set of transcription factors sufficient to induce the anaerobicesponse. The observation that at least one RAP2 gene is highlyxpressed in each cell type (Fig. 2b) supports the role of these genesn the activation of the molecular response to anaerobiosis. Thiss further reinforced by the actual association of these mRNAs toolysomes under aerobic conditions (Fig. 2c). In fact, it suggestshat a constant translation rate provides the accumulation of ERF-II transcription factors when their proteasomal degradation isampered under low oxygen conditions. An additional reservoirf RAP2.12 protein has been shown to be built up at the plasmaembrane [11] and it is possible that the same occurs to the other

RF-VII proteins.The major role played by RAP2.2 and RAP2.12 in the initial acti-

ation of the anaerobic response is supported by the observationhat plants devoid of a functional RAP2.12 gene and in which RAP2.2as significantly down-regulated, such as the double rap2.2-12utants generated in this study, display an extremely reduced

ctivation of four anaerobic marker genes (Fig. 5b). Apparently,he inability to efficiently activate the anaerobic response did notlter normal plant phenotypes, although it cannot be excluded thatAP2.2 expression in these double mutant plants was sufficiento sustain a certain degree of activation of the anaerobic geneshroughout development.

The trans-activation capacity of RAP2.2 and RAP2.12 is associ-ted with their ultimate C-terminal domain (Fig. 3), which includes

conserved LWSY motif previously identified in C-repeat-bindingranscription factors (CBFs). Further analyses aimed at identify-ng proteins able to bind this C-terminal motif will shed newight on the mechanisms by which transcription is activated byAP2.2 and RAP2.12. Usually, transcription factors are believed toodulate gene expression either by directly interacting with the

eneral transcription machinery and their co-activators, therebyffecting complex formation, or by recruiting chromatin remodel-ng complexes at specific genomic regions [27]. Many responses tonvironmental cues are indeed mediated by chromatin modifica-ions and hypoxia, in particular, was shown to affect the acetylationnd methylation level of the H3 histones in octamers located at theDH1 and PDC1 loci of rice seedlings [28]. As PDC1 and ADH areontrolled by RAP2.2 and RAP2.12 in Arabidopsis [11] and [13], itould be interesting to test whether these transcription factors are

ble to control the chromatin state of their target genes.

cknowledgements

This work was financially supported by the Scuola Superioreant’Anna and the Max Planck Institute of Molecular Plant Physiol-gy. Liem T. Bui was supported by the International PhD Programmen Agrobiodiversity (Scuola Superiore Sant’Anna).

ppendix A. Supplementary data

Supplementary data associated with this article can beound, in the online version, at http://dx.doi.org/10.1016/j.lantsci.2015.03.008.

eferences

[1] E. Loreti, A. Poggi, G. Novi, A. Alpi, P. Perata, A genome-wide analysis of the

effects of sucrose on gene expression in Arabidopsis seedlings under anoxia,Plant Physiol. 137 (2005) 1130–1138.

[2] R. Lasanthi-Kudahettige, L. Magneschi, E. Loreti, S. Gonzali, F. Licausi, G. Novi,O. Beretta, F. Vitulli, A. Alpi, P. Perata, Transcript profiling of the anoxic ricecoleoptile, Plant Physiol. 144 (2007) 218–231.

[

e 236 (2015) 37–43 43

[3] C. Branco-Price, R. Kawaguchi, R.B. Ferreira, J. Bailey-Serres, Genome-wide anal-ysis of transcript abundance and translation in Arabidopsis seedlings subjectedto oxygen deprivation, Ann. Bot. 96 (2005) 647–660.

[4] A. Mustroph, S.C. Lee, T. Oosumi, M.E. Zanetti, H. Yang, Cross-kingdom compar-ison of transcriptomic adjustments to low-oxygen stress highlights conservedand plant-specific responses, Plant Physiol. 152 (2010) 1484–1500.

[5] K.P. Ismond, R. Dolferus, M. De Pauw, E.S. Dennis, A.G. Good, Enhanced lowoxygen survival in Arabidopsis through increased metabolic flux in the fer-mentative pathway, Plant Physiol. 132 (2003) 1292–1302.

[6] D.A. Weits, B. Giuntoli, M. Kosmacz, S. Parlanti, H.-M. Hubberten, Plant cysteineoxidases control the oxygen-dependent branch of the N-end-rule pathway, Nat.Commun. 5 (2014) 3425.

[7] B. Giuntoli, S.C. Lee, F. Licausi, M. Kosmacz, T. Oosumi, A trihelix DNA bind-ing protein counterbalances hypoxia responsive transcriptional activation inArabidopsis, PLOS Biol. 9 (2014) e1001950.

[8] R. Chang, C.J. Jang, C. Branco-Price, P. Nghiem, J. Bailey-Serres, Transient MPK6activation in response to oxygen deprivation and reoxygenation is mediatedby mitochondria and aids seedling survival in Arabidopsis, Plant Mol. Biol. 78(2012) 109–122.

[9] C. Pucciariello, S. Parlanti, V. Banti, G. Novi, P. Perata, Reactive oxygen species-driven transcription in Arabidopsis under oxygen deprivation, Plant Physiol.159 (2012) 184–196.

10] D.J. Gibbs, S.C. Lee, I.N. Md, S. Gramuglia, T. Fukao, G.W. Bassel, C.S. Correia,F. Corbineau, F.L. Theodoulou, J. Bailey-Serres, M.J. Holdsworth, Homeostaticresponse to hypoxia is regulated by the N-end rule pathway in plants, Nature479 (2011) 415–418.

11] F. Licausi, M. Kosmacz, D.A. Weits, B. Giuntoli, F.M. Giorgi, L.A.C.J. Voesenek, P.Perata, J.T. van Dongen, Oxygen sensing in plants is mediated by an N-end rulepathway for protein destabilization, Nature 479 (2011) 419–422.

12] D.J. Gibbs, J. Bacardit, A. Bachmair, M.J. Holdsworth, The eukaryotic N-end rulepathway: conserved mechanisms and diverse functions, Trends Cell Biol. 10(2014) 603–611.

13] M. Hinz, I.W. Wilson, J. Yang, K. Buerstenbinder, D. Llewellyn, S.E.S. Dennis,M. Sauter, R. Dolferus, Arabidopsis RAP2.2: An ethylene response transcriptionfactor that is important for hypoxia survival, Plant Physiol. 153 (2010) 757–772.

14] F. Licausi, J.T. van Dongen, B. Giuntoli, G. Novi, A. Santaniello, HRE1 and HRE2,two hypoxia-inducible ethylene response factors, affect anaerobic responsesin Arabidopsis thaliana, Plant J. 62 (2010) 302–315.

15] D.J. Gibbs, N. Md Isa, M. Movahedi, J. Lozano-Juste, G.M. Mendiondo, S. Berck-han, N. Marín-de la Rosa, J.V. Conde, C.S. Correia, S.P. Pearce, G.W. Bassel, B.Hamali, P. Talloji, D.F.A. Tomé, A. Coego, J. Beynon, D. Alabadí, A. Bachmair, J.León, J.E. Gray, F.L. Theodoulou, M.J. Holdsworth, Nitric oxide sensing in plantsis mediated by proteolytic control of group VII ERF transcription factors, Mol.Cell 53 (2014) 369–379.

16] J.T. van Dongen, A. Fröhlich, S.J. Ramírez-Aguilar, N. Schauer, A.R. Fernie1, A.Erban, J. Kopka, J. Clark, A. Langer, P. Geigenberger, Transcript and metaboliteprofiling of the adaptive response to mild decreases in oxygen concentrationin the roots of Arabidopsis plants, Ann. Bot. 103 (2009) 269–280.

17] M. Karimi, D. Inzé, A. Depicker, 2 GATEWAY vectors for Agrobacterium-mediated plant transformation, Trends Plant Sci. 7 (2002) 193–195.

18] F. Licausi, D.A. Weits, B.D. Pant, W.R. Scheible, P. Geigenberger, J.T. van Don-gen, Hypoxia responsive gene expression is mediated by various subsets oftranscription factors andmiRNAs that are determined by the actual oxygenavailability, New Phytol. 190 (2011) 442–456.

19] S.D. Yoo, Y.H. Cho, J. Sheen, Arabidopsis mesophyll protoplasts: a versatile cellsystem for transient gene expression analysis, Nat Protoc. 2 (2007) 1565–1572.

20] M. Kosmacz, S. Parlanti, M. Schwarzländer, F. Kragler, F. Licausi, J.T. DongenVAN, The stability and nuclear localization of the transcription factor RAP2.12are dynamically regulated by oxygen concentration, Plant Cell Environ. (2014),http://dx.doi.org/10.1111/pce.12493.

21] K.J. Livak, T.D. Schmittgen, 8 Analysis of relative gene expression data usingreal-time quantitative PCR and the 2(-Delta DeltaC(T)) method, Methods 25(2001) 402–408.

22] T. Hruz, O. Laule, G. Szabó, F. Wessendorp, S. Bleuler, L. Oertle, P. Widmayer,W. Gruissem, P. Zimmermann, Genevestigator V3: A reference expressiondatabase for the meta-analysis of transcriptomes, Adv. Bioinformatics (2008),ID 420747.

23] D. Winter, B. Vinegar, H. Nahal, R. Ammar, G.V. Wilson, N.J. Provart, Electronicfluorescent pictograph” browser for exploring and analyzing large-scale bio-logical data sets, PLoS One 2 (2007) e718.

24] A. Mustroph, M.E. Zanetti, C.J.H. Jang, H.E. Holtan, P.P. Repetti, D.W. Galbraith,T. Girke, J. Bailey-Serres, Profiling translatomes of discrete cell populationsresolves altered cellular priorities during hypoxia in Arabidopsis, Proc. Natl.Acad. Sci. USA 106 (2009) 18843–18848.

25] T. Nakano, K. Suzuki, T. Fujimura, H. Shinshi, Genome-Wide Analysis of the ERFGene Family in Arabidopsis and Rice, Plant Physiol. 140 (2006) 411–432.

26] J.G. Dubouzet, Y. Sakuma, Y. Ito, M. Kasuga, E.G. Dubouzet, S. Miura, M. Seki,K. Shinozaki, K. Yamaguchi-Shinozaki, OsDREB genes in rice, Oryza sativa L.,encode transcription activators that function in drought-, high-salt- and cold-responsive gene expression, Plant J. 33 (2003) 751–763.

27] J.L. Riechmann, Transcriptional regulation: a genomic overview The

Arabidopsis Book (2002) e0085, http://dx.doi.org/10.1199/tab.0085,http://www.bioone.org/[PMC free article] [PubMed].

28] H. Tsuji, H. Saika, N. Tsutsumi, A. Hirai, M. Nakazono, Dynamic and reversiblechanges in histone H3-Lys4 methylation and H3 acetylation occurring atsubmergence-inducible genes in rice, Plant Cell Physiol. 47 (2006) 995–1003.