Binding affinities and interactions among different heat shock element types and heat shock factors in rice (Oryza sativa L.) Dheeraj Mittal 1 , Yasuaki Enoki 2 , Dhruv Lavania 1 , Amanjot Singh 1 , Hiroshi Sakurai 2 and Anil Grover 1 1 Department of Plant Molecular Biology, University of Delhi South Campus, New Delhi, India 2 Division of Health Sciences, Graduate School of Medical Science, Kanazawa University, Japan Keywords heat shock; heat shock element; heat shock protein; heat shock transcription factor; rice (Oryza sativa) Correspondence A. Grover, Department of Plant Molecular Biology, University of Delhi South Campus, New Delhi 110021, India Fax: +91-11-24115270 Tel: +91-11-24117693/24115097 E-mail: anil.anilgrover@gmail.com (Received 9 May 2011, revised 27 June 2011, accepted 29 June 2011) doi:10.1111/j.1742-4658.2011.08229.x Binding of heat shock factors (Hsfs) to heat shock elements (HSEs) leads to transcriptional regulation of heat shock genes. Genome-wide, 953 rice genes contain perfect-type, 695 genes gap-type and 1584 genes step-type HSE sequences in their 1-kb promoter region. The rice genome contains 13 class A, eight class B and four class C Hsfs (OsHsfs) and has OsHsf26 (which is of variant type) genes. Chemical cross-linking analysis of in vitro synthe- sized OsHsf polypeptides showed formation of homotrimers of OsHsfA2c, OsHsfA9 and OsHsfB4b proteins. Binding analysis of polypeptides with oli- gonucleotide probes containing perfect-, gap-, and step-type HSE sequences showed that OsHsfA2c, OsHsfA9 and OsHsfB4b differentially recognize various model HSEs as a function of varying reaction temperatures. The homomeric form of OsHsfA2c and OsHsfB4b proteins was further noted by the bimolecular fluorescence complementation approach in onion epidermal cells. In yeast two-hybrid assays, OsHsfB4b showed homomeric interaction as well as distinct heteromeric interactions with OsHsfA2a, OsHsfA7, OsH- sfB4c and OsHsf26. Transactivation activity was noted in OsHsfA2c, OsH- sfA2d, OsHsfA9, OsHsfC1a and OsHsfC1b in yeast cells. These differential patterns pertaining to binding with HSEs and protein–protein interactions may have a bearing on the cellular functioning of OsHsfs under a range of different physiological and environmental conditions. Structured digital abstract l HSFA2C binds to HSFA2C by cross-linking study (View interaction) l HSFA2C physically interacts with HSFA2C by bimolecular fluorescence complementation (View interaction) l HSFB4B physically interacts with HSFB4B by bimolecular fluorescence complementation (View interaction) l HSFA2A physically interacts with HSFB4B by two hybrid (View interaction) l HSFB4B binds to HSFB4B by cross-linking study (View interaction) l HSFB4B physically interacts with HSF26 by two hybrid (View interaction) l HSFA9 binds to HSFA9 by cross-linking study (View interaction) l HSFA7 physically interacts with HSFB4B by two hybrid (View interaction) l HSFB4B physically interacts with HSFB4C by two hybrid (View interaction) l HSFB4B physically interacts with HSFB4B by two hybrid (View interaction) Abbreviations 3-AT, 3-amino-1,2,4-triazole; BiFC, bimolecular fluorescence complementation; EGS, ethylglycol bis(succinimidylsuccinate); EMSA, electrophoretic mobility shift assay; HS, heat shock; HSE, heat shock element; Hsf, heat shock transcription factor; Hsp, heat shock protein. 3076 FEBS Journal 278 (2011) 3076–3085 ª 2011 The Authors Journal compilation ª 2011 FEBS Introduction The synthesis of heat shock proteins (Hsps) represents one of the most thoroughly studied induced gene expression systems. Hsp genes are primarily regulated by heat stress, metal stress and developmental cues [1,2]. Hsp transcripts increase massively following heat shock (HS), indicating that expression of Hsps is pro- foundly regulated at the transcriptional level. The abil- ity of HS promoters to sense and respond to heat is mainly due to the presence of consensus sequences called heat shock elements (HSEs). The eukaryotic HSE consensus sequence has been defined by altering units of 5¢-nGAAn-3¢. HSEs are separated into three types: perfect (P), gap (G) and step (S) [3]. P-type HSEs have three inverted repeats in a contiguous array (nTTCnnGAAnnTTC). G-type HSEs have two consec- utive inverted sequences, with the third sequence sepa- rated by 5 bp (nTTCnnGAAn(5 bp)nGAAn). S-type HSEs have 5-bp gaps separating all three modules (nTTCn(5 bp)nTTCn(5 bp)nTTCn). In plants, the importance of HSEs for heat-dependent transcriptional regulation is reflected from experiments on promoter deletions and by the capacity of a synthetic HSE sequence integrated in a truncated CaMV35S promoter to stimulate heat-inducible reporter gene expression in transgenic tobacco plants [4]. Heat shock transcription factors (Hsfs) bind with HSEs, eventually resulting in transcriptional activation of HS genes. Hsfs have been characterized from various plant species [5–15]. Plant Hsf circuitry appears more complex than in yeast or animal systems, as Arabidop- sis, rice and tomato contain over 20 Hsf genes [5,16]. Hsfs have a core structure comprising an N-terminal DNA binding domain, an adjacent oligomerization domain with heptad hydrophobic repeats (HR-A ⁄ B), signal sequences for nuclear localization and export and a C-terminal AHA type activation domain [5]. Based on the sequence homology and domain structure, plant Hsfs have been subdivided into three classes, A, B and C [7]. Hsfs are differentially expressed in response to vari- ous abiotic stresses, in a tissue- and stage-specific man- ner [16,17]. It is suggested that the ‘early’ constitutively expressed Hsf genes become activated immediately upon HS and function as the primary regulators in the cell, while the expression of ‘late’ Hsf genes is enhanced sig- nificantly following HS by early Hsfs [18]. The activation of Hsfs occurs in two stages: (a) the induction of high affinity binding to HS promoters accomplished by trimerization and cooperative interac- tions between Hsf trimers and (b) the exposure of one or more dedicated activator domains [19]. Further, divergence of HSE architecture influences gene- and stress-specific responses: it is suggested that HSE archi- tecture is an important determinant of which Hsf mem- bers are recruited and provides enormous functional diversity in transcriptional regulation of target genes [19,20]. Plant Hsfs can potentially form homo- or het- ero-oligomers resulting in altered nuclear localization as well as enhanced or suppressed transcription [21]. In tomato, constitutively expressed HsfA1 and HS-induc- ible HsfA2 have been shown to form hetero-oligomers for nuclear transport [22], and the hetero-oligomers synergistically induce HS response of genes [23]. In Arabidopsis, homomeric interactions for HsfA1a and HsfA1b as well as heteromeric interaction between HsfA1a and HsfA1b were noted [6]. Both HsfA1a and HsfA1b also make heteromeric interactions with HsfA2; however, synergistic transactivation ability is not observed in the heteromers [13]. In contrast to class A Hsfs, class B HsfB1 and HsfB2b proteins showed ho- momeric interactions but did not interact with each other [6]. The homo-oligomer of Arabidopsis HsfA4 acts as an activator of heat stress, whereas HsfA5 forms hetero-oligomers with HsfA4, thereby interfering in the HsfA4 DNA binding capacity and thus acting as a selective repressor [24]. In addition to Hsf–Hsf inter- actions, a large body of information has been accumu- lated on the importance of HSE structure for the differential induction of Hsp genes [9,25–29]. Rice has 25 Hsf genes [30]. In addition, there is one more entry for OsHsf, namely OsHsf26 (LOC_ Os06g22610), which is a predicted 190 amino acid pro- tein and contains an oligomerization domain but lacks the DNA binding domain, signal sequences for nuclear localization and export and the AHA motif [16]. Considering this entry also as Hsf (though a variant type), it is proposed that overall there are 26 genes encoding for the OsHsf family. Taking cues from the information available from Hsf–HSE systems of other organisms, it is presumed that 26 rice Hsfs may have different specificities for HSEs. In this study, we pro- vide detailed genome-wide analysis of rice HSE types. We further provide data on protein–protein interac- tions, DNA binding characteristics and transactivation activity of selected members of class A, B and C OsHsf proteins. Results Genome-wide distribution of HSEs in rice The rice genome database was searched for genes that show HSE-like configurations (as shown in the upper D. Mittal et al. Heat shock element types and heat shock factors in rice FEBS Journal 278 (2011) 3076–3085 ª 2011 The Authors Journal compilation ª 2011 FEBS 3077 panel in Fig. 1A) in their respective 1-kb upstream promoter region (taking A of ATG as +1). These genes were grouped on the basis of the presence of the specific HSE types, irrespective of the number of hits per sequence and the strand of a gene. In totality, 953 genes contained P-type, 695 genes G-type and 1584 genes S-type HSEs in the rice genome (Fig. 1A). In total 711, 476 and 1368 genes contained exclusively the P-, G- and S-type HSEs, respectively; 59 genes showed both P- and G-type HSEs, 56 genes a combi- nation of P- and S-type HSEs, while G- and S-type HSEs were noted together in 33 genes. Also, 127 genes showed all three types of HSEs (P, G and S) in their promoter region. Overall, a higher abundance of S-type HSE was noted compared with P- and G-type HSEs. Employing HS-induced microarray analysis of rice transcripts (unpublished data; also see [2]), we next analyzed how many of the HS induced rice genes from the microarray profiling contain the above HSE configurations. A total of 880 genes showed HS induced transcript profiling (Fig. 1B). Notably, 143 of the 880 HS induced genes contained HSEs in their 1-kb upstream promoter sequences. Out of 143 HS genes, only 22 genes represented the annotated Hsf ⁄ Hsp types. Thus a subset of the HS induced genes ( 16%) contains the typical P-, G- or S-type HSE configurations in their promoters. In vitro analysis of homo-oligomerization potential of OsHsfs and binding of OsHsfs with HSEs OsHsfA2c, OsHsfA7, OsHsfA9, OsHsfB4b, OsHsfB4c and OsHsfC1b polypeptides were synthesized by in vi- tro transcription ⁄ translation reactions (Fig. 2A). When analyzed for their capacity to undergo homo-oligomer- ization by chemical crosslinking, OsHsfA2c, OsHsfA9 and OsHsfB4b showed homotrimer formation activity (shown by black circles) from their respective mono- meric forms (shown by white circles) (Fig. 2B). The formation of trimers in these cases was noted at three different reaction temperatures tested, i.e. 22, 32 and 37 °C. In the case of OsHsfB4b, trimerization was noted at 22 and 32 °C but was barely visible at 37 °C. In contrast, OsHsfA7, OsHsfB4c and OsHsfC1b did not efficiently form homotrimers in the conditions tested. Electrophoretic mobility shift assay (EMSA) was carried out to assess the DNA binding abilities of these OsHsfs (Fig. 3). First, the EMSA was carried out using 32 P-labeled model 3P-type HSE (Fig. 3A) as a function of four different incubation temperatures, i.e. 12, 22, 32 and 37 °C (Fig. 3B). OsHsfA2c, OsHsfA9 and OsHsfB4b polypeptides formed protein–DNA complexes. Binding activity was low at 12 °C and was enhanced with increasing temperature; OsHsfA2c showed maximum binding at 32 and 37 °C; OsHsfA9 showed efficient binding at 22, 32 and 37 °C; and OsHsfB4b showed maximum binding at 22 °C. In con- trast, the DNA affinity of OsHsfA7, OsHsfB4c and OsHsfC1b polypeptides was not significant. Second, the EMSA was carried out at 22 °C employing four different HSE oligonucleotides, i.e. 4P-, 3P-, G- and S-types (Fig. 3C). OsHsfA2c showed high affinity binding to 4P- and 3P-type HSEs and low affinity binding with G- and S-type HSEs. A slowly migrating OsHsfA2c–HSE4P complex was noted: this complex may contain two OsHsfA2c trimers, because coopera- tive binding of two Hsf trimers to 4P-type HSE is observed in yeast Hsf1, Drosophila Hsf and human P-type: nTTCnnGAAnnTTC G-type: nTTCnnGAAn(5 bp)nGAAn S-type: nTTCn(5 bp)nTTCn(5 bp)nTTCn A 0 400 800 1200 1600 B 0 200 400 600 800 1000 HS inducible genes Genes with HSEs P G S P + G P + S G + S P + G + S P* G* S* Fig. 1. Genome-wide distribution of HSEs in rice. (A) Frequency analysis of perfect-type (P-type), gap-type (G-type) and step-type (S-type) HSEs in the 1-kb promoter region of rice genes. P, G and S indicate classes of genes which contained P-type, G-type or S-type in any combination, in totality. P*, G* and S* indicate clas- ses of genes showing exclusively P-type, G-type and S-type HSEs, respectively. (B) Analysis of genes showing HSEs out of the total genes showing HS-inducible expression. Heat shock element types and heat shock factors in rice D. Mittal et al. 3078 FEBS Journal 278 (2011) 3076–3085 ª 2011 The Authors Journal compilation ª 2011 FEBS Hsf1 [19]. OsHsfA9 showed binding to 4P- and 3P- type HSEs but not to G- and S-type HSEs. OsHsfB4b showed binding to 4P- and 3P-type HSEs with much higher affinity than the binding for G-type HSE. Therefore, three OsHsfs differentially recognize various model HSEs in vitro. Transactivation activity of OsHsfs in yeast cells To test the activator potential of OsHsfs, fusions of the Gal4 DNA binding domain and OsHsfs were expressed in yeast cells containing a Gal4-regulated His3 reporter construct [31]. Among five class A OsH- sfs (OsHsfA2a, OsHsfA2c, OsHsfA2d, OsHsfA7 and OsHsfA9) Gal4 fusions of OsHsfA2c, OsHsfA2d and OsHsfA9 supported yeast cell growth on medium con- taining 3-amino-1,2,4-triazole (3-AT), a competitive inhibitor of His3 protein, which suggests that these proteins function as activators in yeast cells (Fig. 4). In class B and class C members, OsHsfC1a and OsH- sfC1b showed transactivation capacity on medium containing 5 m M 3-AT. OsHsf26, a variant type of OsHsf, lacked transactivation activity. Interactions among OsHsfs in yeast and onion epidermal cells Selected OsHsfs were analyzed for their possible ho- momeric and heteromeric interactions, using yeast two-hybrid assays (Fig. 5). In this assay, OsHsfs fused to the Gal4 DNA binding domain and activation domain were expressed in yeast cells containing a Gal4-regulated lacZ reporter construct, and the inter- actions were scored based on b-galactosidase activity. OsHsfA2a, OsHsfA7, OsHsfB4b, OsHsfB4c and OsH- sf26 were tested for homomeric and heteromeric interactions. Of these, OsHsfB4b showed a clear homomeric interaction. OsHsfB4b also showed hetero- meric interactions with OsHsfA7, OsHsfA2a, OsH- sfB4c and OsHsf26 proteins. The bimolecular fluorescence complementation (BiFC) technique has provided support in indicating that Hsfs show protein–protein interactions. This tech- nique has also been used for visualization of the subcel- lular locations of the interacting proteins in the cell [6,32]. Subsequently, OsHsfA2c and OsHsfB4b were analyzed for their potential to form homomers by the Temp ( o C) EGS HsfA9 22 32 37 HsfA7 22 32 37 HsfC1b 22 32 37 HsfB4c 22 32 37 HsfB4b 22 32 37 HsfA2c 22 32 37 70 20 35 50 25 A7 A2c A9 B4c B4b C1b 100 70 140 35 50 25 100 240 A B Fig. 2. Homo-oligomerization activity of in vitro synthesized OsHsf polypeptides. (A) 35 S-labeled OsHsfA2c (A2c), OsHsfA7 (A7), OsHsfA9 (A9), OsHsfB4b (B4b), OsHsfB4c (B4c) and OsHsfC1b (C1b) polypeptides were subjected to SDS ⁄ PAGE electrophoresis and phosphor-imag- ing. Equivalent amounts of polypeptides were electrophoresed, and the different band intensities were due to the different methionine con- tents. The positions of molecular mass markers are shown on the left in kilodaltons. (B) Labeled polypeptides were incubated at the indicated temperatures, treated (+) or untreated ()) with 1.0 m M EGS for 12 min, and subjected to SDS ⁄ PAGE electrophoresis and phos- phor-imaging. Positions of monomers are indicated by white circles, and lower and higher levels of homotrimers are indicated by gray and black circles, respectively. D. Mittal et al. Heat shock element types and heat shock factors in rice FEBS Journal 278 (2011) 3076–3085 ª 2011 The Authors Journal compilation ª 2011 FEBS 3079 BiFC experiment using onion epidermal cells (Fig. 6). A positive BiFC reaction was noted for both OsHsfA2c and OsHsfB4b proteins, indicating that these proteins interact to produce active reporter yellow fluorescent protein. As the BiFC reaction was clearly noted in nuclei in both cases, it is apparent that homomeric forms of these proteins are localized in nuclei. Discussion This study noted that in the rice genome with an esti- mated size of 67 393 genes, 2830 genes contain at least one of the three HSE-type configurations in their 1-kb upstream promoter region. Overall, 953 genes contain P-type, 695 genes G-type and 1584 genes S-type HSEs. pBD-GAL4-OsHsfA2a pBD-GAL4-OsHsfA2c p BD-GAL4-OsHsfA2d pBD-GAL4-OsHsfA7 pBD-GAL4-OsHsfA9 pBD-GAL4-OsHsfB4b pBD-GAL4-OsHsfB4c pBD-GAL4-OsHsfC1a p BD-GAL4-OsHsfC1b pBD-GAL4-OsHsfC2a pBD-GAL4-OsHsf26 pBD-GAL4 SD-WH SD-WH + 1 m M 3-AT SD-WH + 5 m M 3-AT SD-W Fig. 4. Transactivation activity of OsHsfs. Auxotrophic growth assay on SD-W (syn- thetically defined tryptophan dropout med- ium), SD-WH (synthetically defined tryptophan and histidine dropout medium) and SD-WH + 1 m M 3-AT and WH + 5 mM 3-AT (synthetically defined tryptophan and histidine dropout media with 1 and 5 m M of 3-AT). The lane with pBD-GAL4 represents a negative control. Temp ( o C) Effect of Temperature (probe, HSE3P) HsfA9 12 22 32 37 HsfA7 12 22 32 37 HsfB4b 12 22 32 37 HsfA2c 12 22 32 37 HsfA9 4P 3P G S HSE Specificity (at 22 o C) HsfA2c 4P 3P G S HsfB4b 4P 3P G S HsfA7 4P 3P G S Probe HsfC1b 12 22 32 37 HsfB4c 12 22 32 37 HsfB4c 4P 3P G S HsfC1b 4P 3P G S B C 3P HSE-probe: 4P HSE-probe: S HSE-probe: G HSE-probe: tcgacTTCtaGAAgcTTCcaGAAattagtgctactcga A tcgacTTCtaGAAgcTTCcactaattagtgctactcga tcgacTTCtaGAAgctagcaGAAattagtgctactcga tcgacTTCtactagcTTCcactaatTTCtgctactcga Fig. 3. Binding assay of in vitro synthesized OsHsf polypeptides with HSEs. (A) Nucleo- tide sequences of the four HSE oligonucleo- tides are shown. ‘GAA’ and inverted ‘TTC’ sequences are shown by bold uppercase letters. (B) Binding of OsHsfs with 3P-type HSE was analyzed at various temperatures. Unlabeled polypeptides were incubated with 32 P-labeled 3P-type HSE at the indicated temperatures and subjected to PAGE elec- trophoresis and phosphor-imaging. White and black arrowheads indicate positions of unbound DNA fragments and protein–DNA complexes, respectively. (C) Binding of OsHsfs with various HSE types was ana- lyzed at 22 °C. Unlabeled polypeptides were incubated with 32 P-labeled 4P-, 3P-, G- and S-type HSEs at 22 °C and subjected to PAGE electrophoresis and phosphor-imag- ing. Double arrowheads show a complex containing two OsHsfA2c trimers and 4P- type HSE. Heat shock element types and heat shock factors in rice D. Mittal et al. 3080 FEBS Journal 278 (2011) 3076–3085 ª 2011 The Authors Journal compilation ª 2011 FEBS This study further shows that, as only  16% of HS induced genes contain the canonical HSE types, a major population of the HS induced genes are not associated with canonical HSE types. The order of importance of the bases in the nGAAn repeat of HSE is G 2 >A 3 >A 4 , and some deviations from the canonical HSE types are tolerated in functional HSEs [19]. In yeast, some of the HS induced genes contain HSE-like sequences slightly diverged from nGAAn [33]. Nonetheless, it appears that cis-acting sequences different from the typical HSEs may also be playing a role in HS inducibility. HsfA1a of Arabidopsis has been shown to bind TT-rich sequence and stress responsive elements, in addition to P- and G-type HSEs [34]. Recent observations showed that a novel 9-bp AZC ( L-azetidine-2-carboxylic acid) responsive element works as an alternative Hsf binding sequence in rice [35]. It is a possibility that rice genes that are HS inducible and do not have canonical HSEs may harbor such or other novel cis elements. There are reports suggesting that HSEs are present on non-HS-inducible genes as well [27,35,36]. Similar to these observations, we also noted that 2687 rice genes, which are not HS inducible as per our microarray results, have HSEs in their promoter region. Put together, these observations highlight the inadequacies in our current understanding of the relevance of HSEs in HS response in rice. The formation of the trimeric form of Hsfs is con- sidered important for attaining their high affinity bind- ing to HSEs [19]. Using in vitro crosslinking and yeast two-hybrid assays, we noted that OsHsfA2c, OsHsfA9 and OsHsfB4b form homomers. However, homomer formation activity was lacking for OsHsfA2a, OsH- sfA7, OsHsfB4c, OsHsfC1b and OsHsf26 proteins. BiFC analysis showed that OsHsfA2c and OsHsfB4b form a homomeric state and further showed that homomeric OsHsfA2c and OsHsfB4b forms are clearly localized in the nucleus. In vitro, homotrimerized OsH- sfA2c, OsHsfA9 and OsHsfB4b bound to 3P-type HSE. OsHsfB4b showed maximum trimerization and DNA binding activities at lower temperature than 0 0.5 1 1.5 2 2.5 3 3.5 4 Miller units OsOsHsfA7 + A2a OsHsfA7 + B4b OsHsfA7 + B4c OsHsfA7 + 26* OsHsfA7 + pAD OsHsf A2a + pAD OsHsfB4b + pAD OsHsfB4c + B4c OsHsfB4c + 26* OsHsfB4c + pAD OsHsf26* +pAD OsHsf26* + 26* PC NC OsHsfB4b + B4b OsHsfB4b + B4c OsHsfB4b + 26* OsHsf A2a + A2a OsHsf A2a + B4b OsHsf A2a + B4c OsHsf A2a + 26* OsOsHsfA7 + A7 Fig. 5. Interactions among OsHsfs in yeast two-hybrid assays. Yeast two-hybrid assays showing b-galactosidase activity in YRG2 yeast cells transformed with different con- structs. PC, positive control (pSE1111- ScSNF1 and pSE1112-ScSNF4 transformed YRG2 strain to yield YRG2-pSE1111- ScSNF1+pSE1112-ScSNF4 cells); NC, nega- tive control (pAD + pBD vector transformed YRG2 cells). Respective OsHsfs cloned in pBD + pAD transformed in YRG2 cells were also used as a negative control. OsHsfA2c-YFPN 35S OsHsfA2c YFPN NosT OsHsfA2c-YFPC 35S OsHsfA2c YFPC NosT OsHsfB4b-YFPN 35S OsHsfB4b YFPN NosT OsHsfB4b-YFPC 35S OsHsfB4b YFPC NosT A B Fig. 6. Analysis for homomeric protein–protein interactions and subcellular localization of OsHsfA2c and OsHsfB4b by the BiFC approach. (A) Details of the OsHsfA2c–YFPN + OsHsfA2c–YFPC fusion construct are shown in the upper panel. Onion epidermal cells transformed with OsHsfA2c–YFPN+OsHsfA2c–YFPC fusion construct are shown in the lower panel. (B) Details of the OsH- sfB4b–YFPN + OsHsfB4b–YFPC fusion construct are shown in the upper panel. Onion epidermal cells transformed with the OsH- sfB4b–YFPN + OsHsfB4b–YFPC fusion construct are shown in the lower panel. In both (A) and (B), panels in the middle display bright field images while the right panel shows a merged image. NosT refers to nopaline synthase transcription termination signal. D. Mittal et al. Heat shock element types and heat shock factors in rice FEBS Journal 278 (2011) 3076–3085 ª 2011 The Authors Journal compilation ª 2011 FEBS 3081 maximum activities noted in OsHsfA2c and OsHsfA9. This study reflects the first case of plant Hsfs showing that trimerization and DNA binding activities in dif- ferent members is temperature-dependent to differen- tial extents. The trimerization and DNA binding activities at permissive and cooler temperatures (Figs 2 and 3) indicate the possible role of OsHsfs in unstressed control conditions. It remains to be seen what relevance this temperature-dependent pattern under in vitro conditions has in terms of in vivo physio- logical conditions. When various model HSE types were employed in the EMSA, OsHsfA2c showed low- affinity binding to G- and S-type HSEs; however, OsHsfA9 and OsHsfB4b showed no or little binding to G- and S-type HSEs. Although S-type HSEs appear most prevalent in the promoter region in the rice gen- ome, binding with S-type HSE under in vitro condi- tions is of low affinity with the select group of OsHsfs tested herein. Because lower affinity sites contribute in the binding of transcription factors and gene regula- tion [37], the low affinity binding noted in this study may be relevant for the in vivo functioning of OsHsfs. It was postulated that the property of transactivator function resides in the AHA motifs present at the C-terminus [31]. Class A OsHsfs have AHA while class B and C Hsfs lack these motifs [7]. In rice, transactiva- tion activity has previously been reported for OsHsfA2e protein [10]. We noted that three class A proteins (OsHsfA2c, OsHsfA2d and OsHsfA9) show transactivation activity; however, two class A members (OsHsfA2a and OsHsfA7) containing AHA motifs lack activity. Class B OsHsfB4b and OsHsfB4c proteins lacked transactivation activity. Of the OsHsfC1a, OsHsfC1b and OsHsfC2a class C members tested, OsHsfC1a and OsHsfC1b showed transactivation activity that appeared comparable with class A mem- bers in extent. Weak transactivation potential has also been noted in Arabidopsis class C Hsf [31]. It thus seems that elements apart from AHA motifs can make a contribution to transactivation activity. Phosphoryla- tion is implicated in the activation and inactivation of transactivation potential of human HSF1 [38]. We note multiple putative phosphorylation sites in OsHsfs (Table S1). However, involvement of phosphorylation reaction at these sites remains to be established in response to HS conditions. We have earlier shown that OsHsf26 has an oligo- merization domain but lacks the domain for DNA binding activity [16]. The OsHsf26 transcript is expressed under complex stress combinations involving high and low temperatures coupled with oxidative stress in rice seedlings (D. Mittal and A. Grover, unpublished data). In this study, we have shown that OsHsf26 interacts with OsHsfB4b. It is possible that OsHsf26 works as a competitive inhibitor of OsHsfB4b oligomerization, which results in inhibition of DNA binding of OsHsfB4b. Thus, the variant OsHsf26 form may have regulatory roles controlling downstream gene expression via non-functional hetero-oligomeriza- tion with OsHsfs. From the above account, we note that OsHsfB4b is predominantly involved in rice HS response. OsHsfB4b forms homomeric interactions to form a trimeric state and makes heteromeric interactions with various OsH- sfs. We have recently noted that OsHsfB4b binds to OsHsfA2c and OsClpB-cyt ⁄ Hsp100 protein (A. Singh, D. Mittal, D. Lavania, M. Agarwal, R. C. Mishra and A. Grover, unpublished data). In addition, high tran- script abundance of OsHsfB4b gene is noted following heat stress and oxidative stress [16]. OsHsfB4b itself lacks transactivation ability, implying that OsHsfB4b homotrimer binds to P- and G-type HSEs and represses transcription. It is also possible that HSE specificity of OsHsfB4b changes via hetero-oligomerization with OsHsfA2a, OsHsfA7 and OsHsfB4c. The homotrimer of OsHsfA2c may be a potent HS-inducible activator of genes containing various HSE types (the OsHsfA2c transcripts increase upon HS [16]). OsHsfA9 homotri- mer may be involved in activation of genes containing P-type HSEs; however, its transcripts are relatively low under normal and stress conditions [16]. In conclusion, we suggest that these differential patterns may have a bearing on cellular functioning of OsHsfs under a range of different physiological and environmental conditions, which influence synthesis of different target proteins governed by HSE–Hsf interactions. Materials and methods Genome-wide analysis of HSE in the rice genome The 1-kb upstream regions to the translation start site of all the rice genes were downloaded from the RGA database (http://rice.plantbiology.msu.edu/; release 6.1) and analyzed for the presence of consensus P-, S- and G-type HSE ele- ments (Fig. 1A). The motif search function of the CLC Main Workbench 5 (http://www.clcbio.com) was employed to execute the respective queries in default parameters. Cloning of rice OsHsf genes Rice OsHsf genes were cloned from respective KOME clones (Rice Genome Resource Centre, National Institute of Agrobiological Sciences, Tsukuba, Japan) using HS responsive cDNA synthesized in the laboratory. Various primers used in this study are shown in Table S2. Heat shock element types and heat shock factors in rice D. Mittal et al. 3082 FEBS Journal 278 (2011) 3076–3085 ª 2011 The Authors Journal compilation ª 2011 FEBS In vitro polypeptide synthesis, chemical crosslinking and EMSA OsHsf cDNA fragments were cloned into the pTNT vector (Promega, Madison, MA, USA). Using these plasmid DNAs, OsHsf polypeptides were prepared by in vitro tran- scription ⁄ translation (TNT coupled reticulocyte lysate sys- tem with SP6 RNA polymerase; Promega) according to the manufacturer’s protocol, except that the reaction was carried out at 22 °C for 2.5 h. The relative amounts of polypeptides were normalized by the levels of incorporated [ 35 S] methionine and by the methionine contents. Equal amounts of OsHsf polypeptides were subjected to chemical crosslinking analysis with ethylglycol bis(succinimidylsucci- nate) (EGS) and to EMSA with 32 P-labeled oligonucleotide probes containing 4P-, 3P-, G- and S-type HSE sequences (see Fig. 3A) as described [39], except that the reaction mixtures contained 50 m M NaCl. Transactivation assay in yeast cells OsHsf ORFs PCR amplified using gene-specific primer sets (Table S2) were cloned in the pBD-GAL4 vector (Stratagene Agilent Technologies, La Jolla, CA, USA) in the EcoRI and SmaI sites. All PCRs were done using PhusionÔ Hi-Fi DNA polymerase in the presence of 3% dimethyl sulfoxide. pBD- GAL4+OsHsfs were introduced into yeast strain YRG2 (MATa ura3-52 his3-200 ade2-101 lys2-801 trp1-901 leu2-3 112gal4-542 gal80-538 LYS2::UAS GAL1 -TATA GAL1 -HIS3 URA3::UAS GAL4_17mers_(x3) -TATA CYC1 -lacZ) (Stratagene). Transformants were allowed to grow in SD medium lack- ing amino acid tryptophan at 28 °C. Different dilutions were spotted on SD medium lacking amino acids trypto- phan and histidine. To check leaky expression of HIS3 reporter gene, dilutions were also dotted on the respective dropout media supplemented with 1 m M and 5 mM of 3-AT. Each experiment was repeated three times. Yeast two-hybrid assay Yeast two-hybrid assays were carried out using pAD- GAL4 (activation domain fusion, prey) and pBD-GAL4 (binding domain fusion, bait) vectors (Stratagene). For cloning in pAD-GAL4 vector (Stratagene), OsHsf frag- ments excised from pBD-GAL4+OsHsfs by EcoRI and XbaI digestion were cloned in pAD-GAL4 vector in the sites EcoRI and XbaI. OsHsf-bait and OsHsf-prey pairs were co-transformed into YRG2 and transformants were selected on medium lacking leucine and tryptophan. b-galactosidase activity was measured by the quantitative liquid culture method using O-nitrophenyl b- D-galactopyr- anoside as substrate and by filter lift assay (Yeast Proto- cols Handbook; Clontech Laboratories Inc., Mountain View, CA, USA). Each experiment was repeated three times. BiFC assays PCR amplified OsHsfA2c and OsHsfB4b genes were cloned into BiFC vectors pUC-SPYCE and pUC-SPYNE [32]. For transient expression in onion epidermal cells, the fusion proteins with N- or C-terminal parts of yellow fluorescent protein in pUC-SPYCE and pUC-SPYNE vectors were introduced into onion epidermal cells by particle bombard- ment as described previously [2]. After incubation for 16 h, the cells were visualized by confocal laser scanning micro- scope (Leica TCS SP5). Yellow fluorescent protein was excited with an argon laser at 514 nm. Acknowledgements DM, DL and AS acknowledge the Council of Scien- tific and Industrial Research (CSIR), New Delhi, for the Fellowship award. BiFC vectors pUC-SPYCE and pUC-SPYNE were kindly provided by F. Schoffl and C. Oecking, University of Tubingen, Germany. This work was supported in part by the Indo-Finland pro- ject grant from the Department of Biotechnology (DBT), Government of India, to AG and Grants-in- Aid for Scientific Research from the Ministry of Edu- cation, Culture, Sports, Science and Technology of Japan to HS. References 1 Sarkar NK, Kim YK & Grover A (2009) Rice sHsp genes: genomic organization and expression profiling under stress and development. BMC Genomics 10, 393. 2 Singh A, Singh U, Mittal D & Grover A (2010) Gen- ome-wide analysis of rice ClpB ⁄ HSP100, ClpC and ClpD genes. BMC Genomics 11 , 95. 3 Yamamoto A, Mizukami Y & Sakurai H (2005) Identi- fication of a novel class of target genes and a novel type of binding sequence of heat shock transcription factor in Saccharomyces cerevisiae. J Biol Chem 280, 11911– 11919. 4 Scho ¨ ffl F, Rieping M, Baumann G, Bevan M & Angermu ¨ ller S (1989) The function of plant heat shock promoter elements in the regulated expression of chimaeric genes in transgenic tobacco. Mol Gen Genet 217, 246–253. 5 von Koskull-Do ¨ ring P, Scharf KD & Nover L (2007) The diversity of plant heat stress transcription factors. Trends Plant Sci 12, 452–457. 6 Li M, Doll J, Weckermann K, Oecking C, Berendzen KW & Scho ¨ ffl F (2010) Detection of in vivo inter- actions between Arabiopsis class A-HSFs, using a novel BiFc fragment, and identification of novel class B-HSF interacting proteins. Eur J Cell Biol 89, 126–132. D. Mittal et al. Heat shock element types and heat shock factors in rice FEBS Journal 278 (2011) 3076–3085 ª 2011 The Authors Journal compilation ª 2011 FEBS 3083 7 Nover L, Bharti K, Do ¨ ring P, Mishra SK, Ganguli A & Scharf KD (2001) Arabidopsis and the heat stress transcription factor world: how many heat stress tran- scription factors do we need? Cell Stress Chaperones 6, 177–189. 8 Baniwal SK, Bharti K, Chan KY, Fauth M, Ganguli A, Kotak S, Mishra SK, Nover L, Port M, Scharf KD et al. (2004) Heat stress response in plants: a complex game with chaperones and more than twenty heat stress transcription factors. J Biosci 29, 471–487. 9 Kotak S, Vierling E, Ba ¨ umlein H & von Koskull-Do ¨ r- ing P (2007) A novel transcriptional cascade regulating expression of heat stress proteins during seed develop- ment of Arabidopsis. Plant Cell 19, 182–195. 10 Yokotani N, Ichikawa T, Kondou Y, Matsui M, Hirochika H, Iwabuchi M & Oda K (2008) Expression of rice heat stress transcription factor OsHsfA2e enhances tolerance to environmental stresses in trans- genic Arabidopsis. Planta 227, 957–967. 11 Yoshida T, Sakuma Y, Todaka D, Maruyama K, Qin F, Mizoi J, Kidokoro S, Fujita Y, Shinozaki K & Yamaguchi-Shinozaki K (2008) Functional analysis of an Arabidopsis heat-shock transcription factor HsfA3 in the transcriptional cascade downstream of the DREB2A stress-regulatory system. Biochem Biophys Res Commun 11, 515–521. 12 Liu JG, Qin QL, Zhang Z, Peng RH, Xiong AS & Chen JM (2009) OsHSF7 gene in rice, Oryza sativa L., encodes a transcription factor that functions as a high temperature receptive and responsive factor. BMB Rep 42, 16–21. 13 Li M, Berendzen KW & Scho ¨ ffl F (2010) Promoter specificity and interactions between early and late Ara- bidopsis heat shock factors. Plant Mol Biol 73, 559–567. 14 Nishizawa A, Tainaka H, Yoshida E, Tamoi M, Yabu- ta Y & Shigeoka S (2010) The 26S Proteasome function and Hsp90 activity involved in the regulation of HsfA2 expression in response to oxidative stress. Plant Cell Physiol 51, 486–496. 15 Xin H, Zhang H, Chen L, Li X, Lian Q, Yuan X, Hu X, Cao L, He X & Yi M (2010) Cloning and character- ization of HsfA2 from Lily (Lilium longiflorum). Plant Cell Rep 29, 875–885. 16 Mittal D, Chakrabarti S, Sarkar A, Singh A & Gro- ver A (2009) Heat shock factor gene family in rice: genomic organization and transcript expression profil- ing in response to high temperature, low temperature and oxidative stresses. Plant Physiol Biochem 47, 785–795. 17 Swindell WR, Huebner M & Weber AP (2007) Tran- scriptional profiling of Arabidopsis heat shock proteins and transcription factors reveals extensive overlap between heat and non-heat stress response pathways. BMC Genomics 8, 125. 18 Wunderlich M, Doll J, Busch W, Kleindt CK, Loh- mann C & Scho ¨ ffl F (2007) Heat shock factors: regula- tors of early and late functions in plant stress response. Plant Stress 1, 16–22. 19 Sakurai H & Enoki Y (2010) Novel aspects of heat shock factors: DNA recognition, chromatin modulation and gene expression. FEBS J 277, 4140–4149. 20 Yamamoto N, Takemori Y, Sakurai M, Sugiyama K & Sakurai H (2009) Differential recognition of heat shock elements by members of the heat shock transcription factor family. FEBS J 276, 1962–1974. 21 Miller G & Mittler R (2006) Could heat shock tran- scription factors function as hydrogen peroxide sensors in plants? Ann Bot 98, 279–288. 22 Scharf KD, Heider H, Ho ¨ hfeld I, Lyck R, Schmidt E & Nover L (1998) The tomato Hsf System: HsfA2 needs interaction with HsfA1 for efficient nuclear import and may be localized in cytoplasmic heat stress granules. Mol Cell Biol 18, 2240–2251. 23 Chan-Schaminet KY, Baniwal SK, Bublak D, Nover L & Scharf KD (2009) Specific interaction between tomato HsfA1 and HsfA2 creates hetero-oligomeric su- peractivator complexes for synergistic activation of heat stress gene expression. J Biol Chem 284, 20848–20857. 24 Baniwal SK, Chan KY, Scharf KD & Nover L (2007) Role of heat stress transcription factor HsfA5 as specific repressor of HsfA4. J Biol Chem 282, 3605–3613. 25 Carranco R, Almoguera C & Jordano J (1999) An imperfect heat shock element and different upstream sequences are required for the seed-specific expression of a small heat shock protein gene. Plant Physiol 121, 723–730. 26 Almoguera C, Prieto-Dapena P & Jordano J (1998) Dual regulation of a heat shock promoter during embryogenesis: stage-dependent role of heat shock ele- ments. Plant J 13, 437–446. 27 Diaz-Martin J, Almoguera C, Prieto-Dapena P, Espin- osa JM & Jordano J (2005) Functional interaction between two transcription factors involved in the devel- opmental regulation of a small heat stress protein gene promoter. Plant Physiol 139, 1483–1494. 28 Nishiwaza A, Yabuta Y, Yoshida E, Maruta T, Yoshim- ura K & Shigeoka S (2006) Arabidopsis heat shock tran- scription factor as a key regulator in response to several types of environmental stress. Plant J 48, 535–547. 29 Nishizawa A, Yoshida E, Yabuta Y & Shigeoka S (2009) Analysis of the regulation of target genes by an Arabidopsis heat shock transcription factor, HsfA2. Biosci Biotechnol Biochem 73, 890–895. 30 Guo J, Wu J, Ji Q, Wang C, Luo L, Yuan Y, Wang Y & Wang J (2008) Genome-wide analysis of heat shock transcription factor families in rice and Arabidopsis. J Genet Genomics 35, 105–118. 31 Kotak S, Port M, Ganguli A, Bicker F & von Koskull- Do ¨ ring P (2004) Characterization of C-terminal Heat shock element types and heat shock factors in rice D. Mittal et al. 3084 FEBS Journal 278 (2011) 3076–3085 ª 2011 The Authors Journal compilation ª 2011 FEBS domains of Arabidopsis heat stress transcription factors (Hsfs) and identification of a new signature combina- tion of plant class A Hsfs with AHA and NES motifs essential for activator function and intracellular locali- zation. Plant J 39 , 98–112. 32 Walter M, Chaban C, Schu ¨ tze K, Batistic O, Wecker- mann K, Na ¨ ke C, Blazevic D, Grefen C, Schumacher K, Oecking C et al. (2004) Visualization of protein interactions in living plant cells using bimolecular fluo- rescence complementation. Plant J 40 , 428–438. 33 Sakurai H & Takemori Y (2007) Interaction between heat shock transcription factors (HSFs) and divergent binding sequences: binding specificities of yeast HSFs and human HSF1. J Biol Chem 282, 13334–13341. 34 Guo L, Chen S, Liu K, Liu Y, Ni L, Zhang K & Zhang L (2008) Isolation of heat shock factor HsfA1a-binding sites in vivo revealed variations of heat shock elements in Arabidopsis thaliana. Plant Cell Physiol 49, 1306– 1315. 35 Guan JC, Yeh CH, Lin YP, Ke YT, Chen MT, You JW, Liu YH, Lu CA, Wu SJ & Lin CY (2010) A 9 bp cis-element in the promoters of class I small heat shock protein genes on chromosome 3 in rice mediates L-azetidine-2-carboxylic acid and heat shock responses. J Exp Bot 61, 4249–4261. 36 Hahn JS, Hu Z, Thiele DJ & Iyer VR (2004) Genome wide analysis of the biology of stress responses through heat shock transcription factor. Mol Cell Biol 24, 5249– 5256. 37 Pan Y, Tsai CJ, Ma B & Nussinov R (2010) Mecha- nisms of transcription factor selectivity. Trends Genet 26, 75–83. 38 Guettouche T, Boellmann F, Lane WS & Voellmy R (2005) Analysis of phosphorylation of human heat shock factor 1 in cells experiencing a stress. BMC Bio- chem 6,4. 39 Enoki Y & Sakurai H (2011) Diversity in DNA recogni- tion by heat shock transcription factors (HSFs) from model organisms. FEBS Lett 585, 1293–1298. Supporting information The following supplementary material is available: Table S1. In silico analysis for the putative phosphory- lation sites in rice Hsfs. Table S2. List of primers used in this study. This supplementary material can be found in the online version of this article. Please note: As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be reorganized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors. D. Mittal et al. Heat shock element types and heat shock factors in rice FEBS Journal 278 (2011) 3076–3085 ª 2011 The Authors Journal compilation ª 2011 FEBS 3085 . Binding affinities and interactions among different heat shock element types and heat shock factors in rice (Oryza sativa L. ) Dheeraj Mittal 1 , Yasuaki. indicated by gray and black circles, respectively. D. Mittal et al. Heat shock element types and heat shock factors in rice FEBS Journal 278 (201 1) 3076–3085