he ABRE-binding bZIP transcription factor OsABF2 is a positive regulator ofbiotic stress and ABA signaling in rice
d. Amir Hossaina,1, Jung-Il Chob,c,1, Muho Hanb,c, Chul-Hyun Ahna, Jong-Seong Jeonb,c,ynheung Anc,d, Phun Bum Parka,∗
Department of Bioscience and Biotechnology, University of Suwon, San 2-2 Wauri Bongdameup, Hwasung 445-743, Republic of KoreaGraduate School of Biotechnology, Kyung Hee University, Yongin 446-701, Republic of KoreaCrop Biotech Institute, Kyung Hee University, Yongin 446-701, Republic of KoreaDepartment of Plant Molecular Systems Biotechnology, Kyung Hee University, Yongin 446-701, Republic of Korea
r t i c l e i n f o
rticle history:eceived 26 February 2010eceived in revised form 28 May 2010ccepted 29 May 2010
eywords:biotic stressBA-responsive element
a b s t r a c t
Abscisic acid (ABA) is an important phytohormone involved in abiotic stress tolerance in plants. Thegroup A bZIP transcription factors play important roles in the ABA signaling pathway in Arabidopsis butlittle is known about their functions in rice. In our current study, we have isolated and characterizeda group A bZIP transcription factor in rice, OsABF2 (Oryza sativa ABA-responsive element binding fac-tor 2). It was found to be expressed in various tissues in rice and induced by different types of abioticstress treatments such as drought, salinity, cold, oxidative stress, and ABA. Subcellular localization anal-ysis in maize protoplasts using a GFP fusion vector indicated that OsABF2 is a nuclear protein. In yeast
ZIPsABF2ice
experiments, OsABF2 was shown to bind to ABA-responsive elements (ABREs) and its N-terminal regionfound to be necessary to transactivate a downstream reporter gene. A homozygous T-DNA insertionalmutant of OsABF2 is more sensitive to salinity, drought, and oxidative stress compared with wild typeplants. In addition, this Osabf2 mutant showed a significantly decreased sensitivity to high levels of ABAat germination and post-germination. Collectively, our present results indicate that OsABF2 functions asa transcriptional regulator that modulates the expression of abiotic stress-responsive genes through anABA-dependent pathway.
Abiotic stresses constitute a major constraint for agriculturalroduction worldwide, particularly in developing countries whererop growth and productivity can be significantly reduced byhese stimuli. Improvements in the abiotic stress tolerance capac-ty of plants through biotechnology thus holds great promise forncreasing food production in regions with limited resources but
his requires a better understanding of the molecular mechanismsnderlying abiotic stress adaptation in plants (Xiong et al., 2002;hu, 2002; Yamaguchi-Shinozaki and Shinozaki, 2006; Seki et al.,007).
The phytohormone abscisic acid (ABA) controls various aspectsof plant growth and development (Finkelstein et al., 2002;Himmelbach et al., 2003). During vegetative growth, one of themajor roles of ABA is to mediate adaptive responses to variousenvironmental stresses, such as drought, high salinity, low temper-ature, oxidative stress, and mechanical wounding. The pathwaysleading to stress adaptation are categorized as ABA-dependent andABA-independent. In addition, underlying the ABA-mediated stressresponses is the transcriptional regulation of stress-responsivegene expression (Shinozaki and Yamaguchi-Shinozaki, 2000; Xionget al., 2002; Zhu, 2002; Finkelstein et al., 2002; Jakoby et al., 2002;Himmelbach et al., 2003).
During the response and adaptation of plants to diverse abioticstresses, many stress-related genes are induced and the levels ofa variety of stress resistance-related functional proteins accumu-
late. Numerous genes have also been reported to be upregulatedunder stress conditions in vegetative tissues (Zhu, 2002; Seki etal., 2002). As a trigger of gene expression, transcription factorsplay important regulatory roles in every aspect of the biology ofplants, including growth, development, and the response to abi-
5′-GCATGGACCAGTCAGTGTTCGTCG-3′. The PCR products werethen subcloned into the pENTR/D-TOPO vector (Invitrogen, Carls-
Md.A. Hossain et al. / Journal of P
tic and biotic stress. The basic leucine zipper (bZIP) transcriptionactor family is one of the largest such families in plants and its
embers have diverse roles, particularly in plant stress-responsivend hormone signal transduction (Jakoby et al., 2002; Uno et al.,000; Rodriguez-Uribe and O’Connell, 2006). To activate down-tream gene expression, the bZIP transcription factors interact withBA-responsive elements (ABREs). These are cis-acting elementsontaining the PyACGTGGC core sequence and are present in theromoter region of ABA-inducible genes. Hence, these bZIP tran-cription factors are designated as ABRE-binding factors (ABFs) orBRE-binding proteins (AREBs) (Niu et al., 1999; Kim et al., 1997;amaguchi-Shinozaki and Shinozaki, 2005). In transient experi-ents using Arabidopsis mesophyll protoplasts, the transcription
f a reporter gene driven by ABRE has been demonstrated to bectivated by several ABFs such as AREB1 and AREB2 (Uno et al.,000; Nakashima et al., 2006).
In Arabidopsis, the 75 identified bZIP transcription factors cane classified into 10 subfamilies based on sequence homology.hirteen bZIPs belong to the A group which contains the ABFenes (Jakoby et al., 2002). These bZIP ABF genes, ABF2/AREB1,BF4/AREB2, and ABF3, are upregulated by ABA, dehydration, andalinity stress in vegetative tissues (Choi et al., 2000; Uno et al.,000). The constitutive overexpression of ABF3 in Arabidopsis andice also results in enhanced drought tolerance (Kang et al., 2002;h et al., 2005). Moreover, in rice, overexpression of the positive
egulators of ABA signaling, OsbZIP23 and OsbZIP72, enhances abi-tic stress tolerance (Xiang et al., 2008; Lu et al., 2008) and mutantsf OsABF1, which is a positive regulator of ABA signaling, are moreensitive to drought and salinity (Hossain et al., 2010). Besidesbiotic stress tolerance, ABF proteins also play an important rolen plant growth and development. For example, the Arabidopsisbi5 mutant shows decreased sensitivity to ABA inhibition of seedermination and also an altered ABA-regulated gene expressionrofile, indicating that AtABI5 links ABA signal transduction witheed-specific gene expressions (Finkelstein and Lynch, 2000). Inonocots, a bZIP transcription factor, OsABI5, is involved in rice fer-
ility and stress tolerance (Zou et al., 2008). Furthermore, the AtABI5omologs, TRAB1 and HvABI5 in rice and barley, respectively, phys-
cally interact with their corresponding AtABI3 homologs, OsVP1nd HvVP1, and regulate seed maturation and dormancy by activat-ng ABA-responsive genes (Hobo et al., 1999; Nakamura et al., 2001;asaretto and Ho, 2003). These previous data indicate that ABF pro-eins function in a conserved ABA signal transduction pathway inoth dicot and monocot plant species.
Rice is one of the most important stable crops globally and alsomodel monocot species for molecular research. About 89 bZIP
ranscription factors are predicted in the rice genome belongingo 11 groups based on their DNA-binding specificity and aminocid sequences within the basic and hinge regions. Among these, 14elong to group VI (equivalent to group A in Arabidopsis) which con-ains the ABF genes (Nijhawan et al., 2008). However, despite theiotechnological potential of utilizing the group A AtbZIP orthologs
n rice, relatively few of these factors have been functionally studiedhus far. In our present study, we have identified and functionallyharacterized OsABF2, an abiotic stress-inducible bZIP transcrip-ion factor in rice. The tissue-specific expression of OsABF2 wasxamined and its expression in response to several environmentaltresses and plant hormones was also evaluated. The ABRE-bindingnd transactivation ability of OsABF2 was then tested using a yeastystem. To investigate the in vivo function of OsABF2, a T-DNA inser-ional mutant of this gene was analyzed under salinity, drought,nd oxidative stress conditions. We also compared the ABA sensi-ivity of mutant plants and wild type plants at the germination and
ost-germination stages. Our results suggest that OsABF2 is a pos-
tive regulator of the abiotic stresses response and ABA-dependentignaling transduction pathway in rice.
ysiology 167 (2010) 1512–1520 1513
2. Materials and methods
2.1. Phylogenetic analysis
The ClustalW-EBI program was used for multiple sequencealignments (http://www.ebi.ac.uk/clustalW). The phylogenetictree was constructed using MEGA software version 4.0 via theneighbor-joining method (Tamura et al., 2007). Bootstrap analy-sis was performed with 1000 replicates and bootstrap values areshown as percentages.
2.2. Plant materials, growth conditions and stress treatments
Wild type rice (Oryza sativa L. cultivar Dongjin) seeds were sur-face sterilized with 70% ethanol and immersed in distilled waterfor 1 day at 4 ◦C in the dark and then grown in a plant growthchamber (27 ± 1 ◦C, 80% relative humidity and 14/10 h day/nightphotoperiod). Two-week old seedlings were subjected to differ-ent stress treatments including drought (10% polyethylene glycol(PEG)), salinity (250 mM NaCl), oxidative stress (10 �M MethylViologen (MV)), cold (4 ◦C), and ABA (100 �M) for 0, 3, 6, 12, 24,and 48 h. A T-DNA insertional mutant line of OsABF2 was iden-tified from the rice T-DNA Insertion Sequence Database (Jeonget al., 2006; http://www.postech.ac.kr/life/pfg/risd/index.html). Ahomozygous Osabf2 line was isolated by PCR screening usingOsABF2 gene-specific and T-DNA specific primers.
2.3. cDNA cloning
The full length cDNA sequence of OsABF2 was iso-lated from drought treated shoots of rice seedlings (O.sativa cv. Dongjin) by RT-PCR using the forward primer 5′-AAGCTTATGGAGTTGCCGGCGGATGGG-3′ and the reverse primer5′-GAATTCTCAGCATGGACCAGTCAGTGT-3′. The RT-PCR productswere subsequently inserted into the pLUG-TA vector (iNTRONBiotechnology, Seoul, Korea), and sequenced. The confirmed fulllength cDNA sequence has been deposited into NCBI GenBankunder the accession number GU552783.
2.4. RT-PCR
Total RNAs were isolated from the shoot, root, stem, matureleaves, flag leaves, and panicle of rice plants under normalgrowth condition and also from stress-treated seedling shootsusing Trizol reagent (Gibco-BRL, Grand Island, NY). First strandcDNA was synthesized with 2 �g of purified total RNA usingM-MuluTM Reverse Transcriptase (New England Biolabs, Bev-erly, MA) and an oligo (dT) primer in a total volume of25 �l. The RT reaction was incubated at 37 ◦C overnight. Thegene-specific primers 5′-ATCAAGAACAGGGAGTCCGC-3′ and 5′-GAGCCATCACCATTCACCAA-3′ were used in the RT-PCR and the riceactin1 gene (accession number NP 001051086) was amplified as aninternal control to quantify the relative amounts of cDNA (McElroyet al., 1990).
2.5. Subcellular localization of GFP fusion proteins
The OsABF2 gene was amplified by PCR using the primerpairs: forward primer, 5′-CACCATGGAGTTGCCGGCGGATGGGAGC-3′ and either a reverse primer for an N-terminal GFP fusion,5′-TCAGCATGGACCAGTCAGTGTTCG-3′ or C-terminal GFP fusion,
bad, CA). Validated inserts were then subcloned into the respectivedestination vectors p2FGW7 for N-terminal GFP fusion andp2GWF7 for C-terminal GFP fusion (Karimi et al., 2002), using
R clonase (Invitrogen) according to the manufacturer’s instruc-ions. The resulting fusion constructs (GFP-OsABF2 or OsABF2-GFP)riven by the CaMV35S promoter were delivered into maize mes-phyll protoplasts using a PEG-calcium mediated method (Hwangnd Sheen, 2001; Cho et al., 2009), followed by a 12–24 h incu-ation to allow transient expression. Chlorophyll autofluorescencend OsABF1-RFP were used as chloroplast and nuclear markers,espectively (Hossain et al., 2010). Expression of these fusion con-tructs was monitored using a confocal microscope (LSM 510 META,arl Zeiss, Jena, Germany).
.6. Yeast one-hybrid assay
Yeast one-hybrid experiments were performed according to theethods previously described by Hossain et al. (2010). Briefly, the
YC7-Int/ABRE construct was prepared by inserting a trimer ofhe Em1a element (GGACACGTGGCG) into the SmaI site of pYC7-nt. The entire OsABF2 open reading frame (ORF), partial regionsf OsABF2, the C-terminal bZIP region (OsABF2�N), and the N-erminal region excluding the bZIP domain (OsABF2�C), wereespectively amplified by PCR using appropriate primer pairs. TheCR products were cloned into the pENTR/D-TOPO vector (Invit-ogen) and then subcloned into pDEST-GADT7 to fuse to the GAL4ctivation domain (AD) using LR clonase (Invitrogen). The result-ng constructs, AD:OsABF2, AD:OsABF2�N and AD:OsABF2�C,
ere transformed into the yeast strain YIP carrying the pYC7-nt or pYC7-Int/ABRE plasmid containing LacZ reporter gene. TheGADT7/OsABF1 (AD:OsABF1) (Hossain et al., 2010) construct andGADT7 empty vector were also transformed into both yeast strainss positive and negative controls, respectively. The transformedeast colonies were incubated on synthetic defined (SD) mini-al media lacking uracil and leucine (SD/-Ura-Leu) at 30 ◦C. For
he �-galactosidase assay, O-nitrophenyl �-d galactopyranosideONPG) was used as a substrate and the analysis was carried outs described by Hossain et al. (2010) and expressed as Miller units.
.7. Transactivation activity assay
The transactivation activity assay was performed using theatchmaker GAL4 Two-Hybrid System (Clontech, Mountain View,
A). The full length OsABF2, OsABF2�C and OsABF2 �N cloned intohe pENTR/D-TOPO vector were subcloned into pDEST-GBKT7 touse to the GAL4 DNA-binding domain (BD). All constructed vectors,D:OsABF2, BD:OsABF2�N and BD:OsABF2�C, were transformed
nto the yeast strain AH109, respectively. The pGBKT7/OsABF1BD:OsABF1) and pGBKT7 empty vector were also transformednto the yeast strain as positive and negative controls, respec-ively. The transformed yeast colonies were adjusted to OD600 of.5, diluted and dropped on either SD minimal media lacking tryp-ophan (SD/-Trp) or SD minimal media lacking tryptophan andistidine (SD/-Trp-His). The �-galactosidase assay was carried outs described by Hossain et al. (2010).
.8. Stress tolerance in rice mutants
For high salt treatments, rice plants were grown hydroponi-ally in the growth chamber for 2 weeks and then transferred into250 mM NaCl solution for 3 days. After this stress treatment,
lants were transferred to normal hydrophonic growth conditionsor an additional 12 days. For dehydration treatment, 2-week oldice plants were removed from the hydrophonic growth chamber
nd transferred to a dish for 12 h. The plants were then rehydratednd grown for an additional 2 weeks. The number of plants thaturvived and continued to grow was then counted. To generatexidative stress, shoots from 14 days plants were detached fromhe seedlings and floated on 10 �M MV (Catalog no. M2254; Sigma,
hysiology 167 (2010) 1512–1520
St Louis, MO) solution for 64 h in the light. The chlorophyll contentsof the detached shoots with or without exposure to this stress werethen measured according to the method of Ni et al. (2009).
2.9. ABA sensitivity test
To perform the ABA sensitivity test at the germination stage,seed surfaces were sterilized and immersed in distilled water,transferred to the growth chamber, and grown in distilled watersupplemented with different concentration of ABA for 7 days. Thenumber of seeds that successfully germinated was then counted.For post-germination assays, seeds germinated in distilled waterwere transferred to distilled water supplemented with differentconcentrations of ABA for 7 days and the root lengths were mea-sured.
2.10. Statistical analysis
Student’s t-test was performed to determine the significance ofdifferences in survival rate under salinity and dehydration stressesand in chlorophyll content after oxidative stress as well as in rootlength between wild type and homozygous mutants of Osabf2.The online tool available at http://www.physics.csbsju.edu/stats/t-test bulk form.html was used to perform the analysis. Mean andstandard error were calculated by using MS Excel.
3. Results and discussion
3.1. Isolation of the bZIP transcription factor OsABF2
We performed microarray screens for genes which are dif-ferentially regulated under abiotic stress conditions in rice (datanot shown). From these experiments, a bZIP transcription fac-tor, OsABF2, was identified and functionally analyzed. OsABF2was found to be expressed in various tissues and under differ-ent types of abiotic stress. A 975 bp cDNA clone predicted byTIGR Rice Genome Annotation database (http://blast.jcvi.org/euk-blast/index.cgi?project=osa1; LOC Os06g10880) was also isolatedfrom the rice seedling shoots by RT-PCR. From subsequent sequenceanalyses, we found that OsABF2 is identical to OsbZIP46 of the pre-dicted OsbZIP family in rice genome (Nijhawan et al., 2008). Aminoacid sequence analysis further suggested that OsABF2 belongs tothe group A bZIP proteins in rice and that it has all of the typi-cal features of the members of this group, i.e. a bZIP domain andthree conserved domains that are predicted phosphorylation sitesinvolved in stress or ABA signaling and high homology to OsbZIP23followed by OsbZIP72, TRAB1 and ABF2 (Fig. 1). The basic region ofOsABF2 also exhibits a high sequence similarity to the basic domainof these four bZIP proteins. In addition, the leucine zipper regionof OsABF2 contains three heptad repeats. Phylogenetic analysis ofthe group A bZIP transcription factors in both rice and Arabidopsisrevealed that they can be classified into three subgroups (Fig. 2).Interestingly, four of the reported rice bZIPs (OsbZIP23, OsbZIP72,TRAB1 and OsABI5) and five identified Arabidopsis bZIPs (ABF1-4and ABI5) were grouped with OsABF2 in subgroup-III. This resultsuggests the possibility that there is functional conservation amongthese ABFs in rice and Arabidopsis.
3.2. Expression of OsABF2 under abiotic stress conditions
We investigated the tissue-specific expression pattern of OsABF2by RT-PCR and detected transcripts in all of the tissues tested, i.e.the shoot, root, stem, mature leave, flag leave, and panicle, undernormal growth conditions (Fig. 3A). To further examine the physi-ological and functional role of OsABF2, we evaluated its expression
Md.A. Hossain et al. / Journal of Plant Physiology 167 (2010) 1512–1520 1515
F her bZ( P 849d unde
pidO
ig. 1. Multiple sequence alignment of the OsABF2 protein in comparison with otNP 001048225), OsbZIP72 (NP 001063362), TRAB1 (NP 0010622018) and ABF2 (Nouble underline represents the basic region and leucine repeat whereas the single
rofile in shoots under different abiotic stress conditions includ-ng drought, salinity, cold, oxidative stresses, and ABA (Fig. 3). Inrought-, salinity- and oxidative stress- (MV) treated seedlings,sABF2 expression was highly induced within 3 h (Fig. 3B–D). When
IP proteins. An amino acid alignment of OsABF2 with the bZIP proteins, OsbZIP23777) is shown. Blast analysis was performed using the NCBI-P blast program. Therlines indicate the conserved regions.
the seedlings were subjected to cold conditions (4 ◦C), OsABF2expression was induced within 3 h and slightly decreased after24 h (Fig. 3E). Upon exposure to ABA, OsABF2 was again inducedwithin 3 h, reached a maximum level after 12–24 h and then grad-
1516 Md.A. Hossain et al. / Journal of Plant Physiology 167 (2010) 1512–1520
Fig. 2. Phylogenetic analysis of the OsABF2 protein with the group A bZIP proteins from Aras percentages. Bootstrap analysis was performed with 1000 replicates and bootstrap val
Fig. 3. Expression analysis of the OsABF2 gene in various rice tissues under normalconditions and in shoots under different stress conditions. (A) Expression pattern ofthe OsABF2 gene in various rice tissues (Sh-Shoot, S-Stem, R-Root, ML-Mature Leave,FL-Flag Leave, and P-Panicle) under normal conditions. (B-F) Expression pattern ofthe OsABF2 gene under conditions of drought (10% PEG), salinity (250 mM NaCl),oxidative stress (10 �M MV), cold (4 ◦C), and following ABA (100 �M) treatment ofthe shoots of rice seedlings at 0, 3, 6, 12, 24 and 48 h after stress onset. The rice actin1gene was used as an internal control. Similar results were obtained from three repeatexperiments.
abidopsis and rice. The numbers in the branches are the bootstrap values expressedues are shown as percentages.
ually decreased (Fig. 3F). These results demonstrate that OsABF2 israpidly upregulated under all of the abiotic stress conditions tested.
ABF1 to 4 and OsbZIP23, OsbZIP72, OsABI5 and OsABF1 were alsostrongly induced by drought, salt, cold, and ABA treatments andhave been characterized previously for their critical roles in ABA-dependent signaling and stress tolerance (Choi et al., 2000; Kang etal., 2002; Kim et al., 2004a; Xiang et al., 2008; Lu et al., 2008; Zou etal., 2008; Hossain et al., 2010). The strong induction of OsABF2 underabiotic stress and following ABA treatment, and the high similar-ity of OsABF2 to OsbZIP23, OsbZIP72, and ABF1 to 4 suggest thatOsABF2 may play an important role in ABA-dependant signalingand stress tolerance in rice.
3.3. Subcellular localization of OsABF2
We generated GFP-OsABF2 and OsABF2-GFP fusion constructsunder the control of the CaMV35S promoter to determine the sub-cellular localization of the OsABF2 protein. These constructs werethen expressed in the maize mesophyll protoplasts. The resultsshowed that both the OsABF2-GFP and GFP-OsABF2 fusion pro-teins are expressed in the nuclei of maize protoplasts (Fig. 4). Thiscolocalization was confirmed with the nuclear marker, OsABF1-RFP(Hossain et al., 2010). These data indicate that OsABF2 is a nuclearprotein and may thus have a role as a transcription factor.
3.4. DNA-binding and transactivation ability of OsABF2
A yeast one-hybrid experiment was performed to determinethe binding ability of OsABF2. The full length OsABF2, OsABF2�Cand OsABF2�N were cloned into a yeast expression vector
pDEST-GADT7 which harbors the GAL4 activation domain (AD)under the control of the yeast ADH1 promoter. The resultingconstructs, AD:OsABF2, AD:OsABF2�N and AD:OsABF2�C, weretransformed into the yeast strain YIP carrying pYC7-Int plasmidthat contains a trimer of the Em1a element, a conserved ABRE
Md.A. Hossain et al. / Journal of Plant Physiology 167 (2010) 1512–1520 1517
F ansfeca resped e refera
darsaiws
Fsop
ig. 4. Subcellular localization of GFP-OsABF2 and OsABF2-GFP fusion proteins in trutofluorescence and OsABF1-RFP were used as chloroplast and nuclear markers,istinguish it from GFP (green) and RFP (red) fluorescence. (For interpretation of thrticle.)
omain and a lacZ reporter gene. The results of �-galactosidasectivity demonstrated that OsABF2, especially the C-terminalegion, binds to the cis-acting element containing the ABRE coreequence (Fig. 5A). The AD:OsABF1 construct as a positive control
lso showed high �-galactosidase activity. No enzymatic activ-ty was detected in the pYC7-Int plasmid-carrying YIP strain
hich does not contain this conserved ABRE domain (data nothown).
ig. 5. DNA-binding and transactivation activity assays of the OsABF2 protein. (A) ABREystem. AD:OsABF1 and pGADT7 empty vector were used as a positive and a negative contf DNA-binding ability using �-galactosidase activity in yeast. (B) Transactivation activityositive and a negative control, respectively. �-Galactosidase activity is expressed in Mill
ted mesophyll protoplasts of maize, (A) GFP-OsABF1; (B) OsABF1-GFP. Chlorophyllctively. A false color (blue) was used to monitor chlorophyll autofluorescence toences to color in this figure legend, the reader is referred to the web version of the
To examine transactivation ability of OsABF2, the full lengthOsABF2, and also OsABF2�C and OsABF2�N sequences were clonedinto a yeast vector pDEST-GBKT7 which harbors the GAL4 DNA-binding domain (BD). The resulting BD:OsABF2, BD:OsABF2�N and
BD:OsABF2�C constructs were transformed into the yeast strainAH109. The BD:ABF2 transformant grew well on SD/-Trp-His plates.This transactivation ability was also observed for BD:OsABF2�Cbut not in the yeast strain containing BD:OsABF2�N, indicating
-mediated DNA-binding assay of the OsABF2 protein using the yeast one-hybridrol, respectively. Top, vectors used in yeast experiments. Bottom, the measurementassay of the OsABF2 protein. BD:OsABF1 and pGBKT7 empty vector were used as aer units. The bar indicates the standard error.
1518 Md.A. Hossain et al. / Journal of Plant Physiology 167 (2010) 1512–1520
Fig. 6. Survival rate of Osabf2 mutant plants under abiotic stress conditions. (A) Schematic diagram of the OsABF2 genomic structure and T-DNA insertion site in the Osabf2mutant. The box and solid lines indicate exons and introns, respectively. The position of the T-DNA is indicated by a triangle (left). RT-PCR analysis of OsABF2 in the wild typeDongjin (DJ) rice plant, and in the corresponding T-DNA Osabf2 mutant at 12 h after 100 �M ABA treatment (right). (B) Survival rate of Osabf2 mutant plants subjected to highsalinity. The error bars indicate the standard error (triplicates, n = 25 each). (C) Representative phenotype of wild type and Osabf2 mutant plants under high salinity. Plants onthe left and right, respectively, are surviving and dead wild type plants. (D) Survival rate of Osabf2 mutant plants subjected to dehydration. The error bars indicate the standarde f2 mur nce pt ard erl een th
t(tgPOa(2
3
cmT
rror (triplicates, n = 25 each). (E) Representative phenotype of wild type and Osabepresentative surviving and dead plants from each line. (F) Oxidative stress tolerahen measured with and without stress conditions. The error bars indicate the standine and the wild (DJ) are significant (P < 0.05). ** Indicate that the differences betw
hat the N-terminal region of OsABF2 is essential for transactivationFig. 5B). Our current result thus indicates that OsABF2 may func-ion as a transcriptional activator to regulate specific downstreamenes and that this activity is dependent on its N-terminal region.revious yeast experiments with OsbZIP23, OsbZIP72, OsABI5 andsABF1 have also shown that these proteins have transactivationctivity for which their respective N-terminal regions are requiredXiang et al., 2008; Lu et al., 2008; Zou et al., 2008; Hossain et al.,010).
.5. Phenotype analysis of an Osabf2 mutant
To elucidate the in vivo function of OsABF2 under abiotic stressonditions, a loss of function approach was employed. The Osabf2utant generated from Dongjin (DJ) wild type plant contains a
-DNA insertion at 53 bp upstream of the start codon. The abol-
tant plants under dehydration stress. Plants on the left and right, respectively, arehotographs were taken before and after treatment. (G) Chlorophyll contents wereror (triplicates, n = 10 each). * Indicates that the differences between the transgenice transgenic line (Osabf2) and the wild (DJ) are highly significant (P < 0.01).
ishment of OsABF2 transcripts in the homozygous mutants wasconfirmed by RT-PCR (Fig. 6A). To evaluate salt tolerance at theseedling stage, the mutant and corresponding wild type plants weregrown hydroponically for 2 weeks and transferred into a 250 mMNaCl solution for 3 days. After salt treatment, these plants wereallowed to recover in hydroponic solution for a further 12 days.The results showed that 24.5% of the wild type plants continued togrow at the end of the experimental period but all of the mutantplants died after this treatment (Fig. 6B and C). Osabf2 mutant plantsare thus more susceptible to high salinity than wild type plants.
To analyze the drought stress tolerance response of Osabf2
mutant and wild type rice plants at the seedling stage, we con-ducted parallel stress tolerance assays for dehydration. Two-weekold seedlings were exposed to dry conditions for 12 h and thengrown to recovery in a hydroponic solution for more than 2 weeks.The results demonstrated that 30% of the mutant plants and 56%
Md.A. Hossain et al. / Journal of Plant Physiology 167 (2010) 1512–1520 1519
F n. (A)G riplicb ces bet nifica
optmtaDalQiwOw
ttstaap(p
3
ardA
ig. 7. ABA sensitivity of the Osabf2 mutant at germination and post-germinatioermination rate corresponding to (A). The error bars indicate the standard error (tars indicate the standard error (triplicates; n = 6 each). * Indicates that the differenhe differences between the transgenic line (Osabf2) and the wild (DJ) are highly sig
f the wild type plants had survived (Fig. 6D and E). Osabf2 mutantlants were thus found to be less tolerant to dehydration than wildype. We also determined the oxidative stress tolerance of Osabf2
utant plants by evaluating the damage induced by MV. It is knownhat MV causes chlorophyll degradation and cell membrane leak-ge by generating reactive oxygen species (Kurepa et al., 1998).etached shoots were placed in 10 �M MV solution and any dam-ge induced by this treatment was observed visually. The mutanteaves became more bleached compared with wild type (Fig. 6F).uantitative determinations of their chlorophyll contents further
ndicated that the mutant plants lost more chlorophyll than theirild type counterparts (Fig. 6G). These findings indicated that thesabf2 mutant rice plant is more susceptible to oxidative stress thanild type.
Our present results are consistent with previous findings thathe OsbZIP23 mutant is more sensitive to salinity and drought andhat the overexpression of OsbZIP23 enhances tolerance to thesetresses (Xiang et al., 2008), that Osabf1 mutants are more sensitiveo salinity and dehydration in rice (Hossain et al., 2010), that thebf3 and abf4 mutants display defects in response to ABA, salinitynd dehydration in Arabidopsis, and that ABF2 and ABF3 overex-ressing transgenic plants are more tolerant to oxidative stressKim et al., 2004a,b). It is thus highly probable that OsABF2 plays aositive role in abiotic stress signaling.
.6. Decreased ABA sensitivity of Osabf2 mutant
Since OsABF2 was found to be strongly induced by ABA in ournalysis (Fig. 3E), we tested whether its protein product plays aole in the ABA sensitivity of rice, an important aspect of ABA-ependent regulation. The mutant plants were examined for theirBA sensitivity profiles at the germination stage. In seed germina-
Germination performance of the Osabf2 mutant and wild type Dongjin (DJ). (B)ates; n = 22 each). (C) ABA sensitivity analysis of primary root elongation. The errortween the wild (DJ) and the transgenic line are significant (P < 0.05). ** Indicate thatnt (P < 0.01).
tion experiments in which the nutrients were supplemented withABA, the OsABF2 mutant plants had relatively higher germinationrates (84%, 76% and 42% with 3, 5 and 10 �M ABA, respectively)than wild type (74%, 68% and 28% with 3, 5 and 10 �M ABA, respec-tively). Under normal conditions, the germination rates of mutantand wild type plants were equivalent (Fig. 7A and B). These resultsindicate that the OsABF2 mutant is less sensitive to ABA, comparedwith wild type plants and are thus consistent with earlier findingsfor Osbzip23, abi5, abf3, and abf4, mutants which are less sensitive toABA than wild type plants and for which the corresponding overex-pression lines are hypersensitive to ABA (Xiang et al., 2008; Zou etal., 2007; Kim et al., 2004b). We also evaluated the ABA sensitivity ofthe mutant and wild type seedlings post-germination. The resultsshowed that the primary root growth of wild type plants is moresensitive to ABA compared with OsABF2 mutant plants (Fig. 7C).Moreover, shoot growth inhibition by ABA is more severe in wildtype plants (data not shown). These results are consistent withthose obtained for OsbZIP23 and OsbZIP72, in which the mutantOsbZIP23 plant is less sensitive to ABA at the post-germinationstage (Xiang et al., 2008) and the overexpression of OsbZIP23 andOsbZIP72 in transgenic plants results in increased sensitivity to ABAat both the germination and post-germination stages (Xiang et al.,2008; Lu et al., 2008). These results strongly suggest that OsABF2plays a critical role in mediating ABA sensitivity in rice.
The results of our current study show that OsABF2 is expressedin various rice tissues and induced by different abiotic stress stim-uli. The ABRE-mediated DNA-binding and transactivation ability of
OsABF2 was also demonstrated in yeast experiments. The positiveregulatory role of OsABF2 in stress tolerance and ABA sensitivitywas then examined via a loss of function approach. Our presentresults indicate that the OsABF2 gene has significant potential toimprove both stress tolerance and ABA signaling in rice.
1 lant P
A
vTK(vCE
R
C
C
C
F
F
H
H
H
H
J
J
K
K
K
K
K
K
520 Md.A. Hossain et al. / Journal of P
cknowledgments
We thank Professor Soo Young Kim of Chonnam National Uni-ersity, Kwangju, Korea for generously supplying the yeast strains.his work was supported by a Research Foundation Grant of theorean Government (MOEHRD, Basic Research Promotion Fund)
KRF-2005-070-C00128), and by grants from the World Class Uni-ersity (R33-2008-000-10168-0) and Crop Functional Genomicsenter (CG2111-2) programs funded by the Korean Ministry ofducation, Science, and Technology.
eferences
asaretto J, Ho TD. The transcription factors HvABI5 and HvVP1 are required for theabscisic acid induction of gene expression in barley aleurone cells. Plant Cell2003;15:271–84.
ho JI, Ryoo N, Eom JS, Lee DW, Kim HB, Jeong SW, et al. Role of the rice hexokinasesOsHXK5 and OsHXK6 as glucose sensors. Plant Physiol 2009;149:745–59.
hoi H, Hong J, Ha J, Kang J, Kim SY. ABFs, a family of ABA-responsive element bindingfactors. J Biol Chem 2000;275:1723–30.
inkelstein RR, Gampala SS, Rock CD. Abscisic acid signaling in seeds and seedlings.Plant Cell (Suppl) 2002;14:S15–45.
inkelstein RR, Lynch T. The Arabidopsis abscisic acid response gene ABI5 encodes abasic leucine zipper transcription factor. Plant Cell 2000;12:599–609.
immelbach A, Yang Y, Grill E. Relay and control of abscisic acid signaling. Curr OpinPlant Biol 2003;6:470–9.
obo T, Kowyama Y, Hattori T. A bZIP factor, TRAB1, interacts with VP1and mediatesabscisic acid-induced transcription. Proc Natl Acad Sci USA 1999;96:15348–53.
ossain MA, Lee Y, Cho JI, Ahn CH, Lee SK, Jeon JS, et al. The bZIP transcription factorOsABF1 is an ABA responsive element binding factor that enhances abiotic stresssignaling in rice. Plant Mol Biol 2010;72:557–66.
wang I, Sheen J. Two-component circuitry in Arabidopsis cytokinin signal trans-duction. Nature 2001;413:383–9.
akoby M, Weisshaar B, Droge-Laser W, Vicente-Carbajosa J, Tiedemann J, Kroj T, etal. bZIP transcription factors in Arabidopsis. Trends Plant Sci 2002;7:106–11.
eong DH, An S, Park S, Kang HG, Park GG, Kim SR, et al. Generation of a flank-ing sequence-tag database for activation-tagging lines in japonica rice. Plant J2006;45:123–32.
ang JY, Choi HI, Im MY, Kim SY. Arabidopsis basic leucine zipper proteins thatmediate stress-responsive abscisic acid signaling. Plant Cell 2002;14:343–57.
arimi M, Inze D, Depicker A. Gateway vectors for Agrobacterium-mediated planttransformation. Trends Plant Sci 2002;7:193–5.
im JB, Kang JY, Kim SY. Over-expression of a transcription factor regulating ABAresponsive gene expression confers multiple stress tolerance. Plant BiotechnolJ 2004a;2:459–66.
im S, Kang J, Cho D, Park JH, Kim SY. ABF2, an ABRE-binding bZIP factor, is anessential component of glucose signaling and its overexpression affects multiplestress tolerance. Plant J 2004b;40:75–87.
im SY, Chung HJ, Thomas TL. Isolation of a novel class of bZIP transcription factorthat interact with ABA-responsive and embryo-specification elements in the Dc3promoter using a modified yeast one-hybrid system. Plant J 1997;11:1237–51.
urepa J, Smalle J, Montagu MV, Inze D. Oxidative stress tolerance and longevity inArabidopsis: the late flowering mutant gigantea is tolerant to paraquat. Plant J1998;14:759–64.
hysiology 167 (2010) 1512–1520
Lu G, Gao C, Zhong X, Han B. Identification of OsbZIP72 as a positive regulator of ABAresponse and drought tolerance in rice. Planta 2008;229:605–15.
McElroy D, Zhang W, Cao J, Wu R. Isolation of an efficient actin promoter for use inrice transformation. Plant Cell 1990;2:163–71.
Nakamura S, Lynch TJ, Finkelstein RR. Physical interactions between ABA responseloci of Arabidopsis. Plant J 2001;26:627–35.
Nakashima K, Fujita Y, Katsura K, Maruyama K, Narusaka Y, Seki M, et al. Tran-scriptional regulation of ABI3 and ABA responsive genes including RD29B andRD29A in seeds, germinating embryos and seedlings of Arabidopsis. Plant MolBiol 2006;60:51–68.
Ni Z, Kim ED, Chen ZJ. Chlorophyll and Strach Assaays. Nat Protocols Plant Bioldoi:10.1038/nprot.2009.12.
Nijhawan A, Jain M, Tyagi AK, Khurana JP. Genomic survey and gene expression anal-ysis of the basic leucine zipper transcription factor family in rice. Plant Physiol2008;146:333–50.
Niu X, Renshaw-Gegg L, Miller L, Guiltinan MJ. Bipartite determinants of DNA-binding specificity of plant basic leucine zipper proteins. Plant Mol Biol1999;41:1–13.
Oh SJ, Song SI, Kim YS, Jang HJ, Kim SY, Kim M, et al. Arabidopsis CBF3/DREB1A andABF3 in transgenic rice increased tolerance to abiotic stress without stuntinggrowth. Plant Physiol 2005;138:341–51.
Rodriguez-Uribe L, O’Connell MA. A root-specific bZIP transcription factor is respon-sive to water deficit stress in tepary bean (Phaseolus acutifolius) and commonbean (P. vulgaris). J Exp Bot 2006;57:1391–8.
Seki M, Narusaka M, Ishida J, Nanjo T, Fujita M, Oono Y, et al. Monitoring the expres-sion profiles of 7000 Arabidopsis genes under drought, cold and high salinitystresses using a full-length cDNA microarray. Plant J 2002;31:279–92.
Seki M, Urano T, Shinozaki K. Regulatory metabolic networks in drought stressresponses. Curr Opin Plant Biol 2007;10:296–302.
Shinozaki K, Yamaguchi-Shinozaki K. Molecular responses to dehydration and lowtemperature: differences and cross-talk between two stress signaling pathways.Curr Opin Plant Biol 2000;3:217–23.
Tamura K, Dudley J, Nei M, Kumar S, MEGA4:. Molecular Evolutionary GeneticsAnalysis (MEGA) software version 4.0. Mol Biol Evol 2007;24:1596–9.
Uno Y, Furihata T, Abe H, Yoshida R, Shinozaki K, Yamaguchi-Shinozaki K. Ara-bidopsis basic leucine zipper transcription factors involved in an abscisicacid-dependent signal transduction pathway under drought and high-salinityconditions. Proc Natl Acad Sci USA 2000;97:11632–7.
Xiang Y, Tang N, Du H, Ye H, Xiong L. Characterization of OsbZIP23 as a keyplayer of basic leucine zipper transcription factor family for conferring Abscisicacid sensitivity and salinity and drought tolerance in rice. Plant Physiol2008;148:1938–52.
Xiong L, Schumaker KS, Zhu JK. Cell signaling during cold, drought, and salt stress.Plant Cell (Suppl) 2002;14:S165–83.
Yamaguchi-Shinozaki K, Shinozaki K. Organization of cis-acting regulatory ele-ments in osmotic and cold-stress-responsive promoters. Trends Plant Sci2005;10:88–94.
Yamaguchi-Shinozaki K, Shinozaki K. Transcriptional regulatory networks in cellularresponses and tolerance to dehydration and cold stresses. Annu Rev Plant Biol2006;57:781–803.
Zhu JK. Salt and drought stress signal transduction in plants. Annu Rev Plant Biol
2002;53:247–73.
Zou M, Guan Y, Ren H, Zhang F, Chen F. A bZIP transcription factor, OsABI5, is involvedin rice fertility and stress tolerance. Plant Mol Biol 2008;66:675–83.
Zou M, Guan Y, Ren H, Zhang F, Chen F. Characterization of alternative splicingproducts of bZIP transcription factors OsABI5. Biochem Biophys Res Commun2007;360:307–13.
