HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 46, 43866-43872, November 15, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/46/43866    most recent M207330200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Almoguera, C. Articles by Jordano, J. Search for Related Content PubMed PubMed Citation Articles by Almoguera, C. Articles by Jordano, J. Social Bookmarking                      What's this?

A Seed-specific Heat-shock Transcription Factor Involved in Developmental Regulation during Embryogenesis in Sunflower^*

Concepción Almoguera, Anabel Rojas^§, Juan Díaz-Martín^§, Pilar Prieto-Dapena, Raúl Carranco^¶, and Juan Jordano

From the Instituto de Recursos Naturales y Agrobiología, C.S.I.C. Apartado 1052, 41080 Sevilla, Spain

Received for publication, July 22, 2002, and in revised form, September 10, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

We report the cloning and functional characterization of the first heat-shock transcription factor that is specifically expressed during embryogenesis in the absence of environmental stress. In sunflower embryos this factor, HaHSFA9, trans-activated promoters with poor consensus heat-shock cis-elements, including that of the seed-specific Hahsp17.6G1 gene. Mutations that improved the heat-shock cis-element consensus at the Hahsp17.7G4 promoter impaired transient activation by HaHSFA9 in sunflower embryos. The same mutations did not affect heat-shock-induced gene expression of this promoter in transgenic tobacco plants but reduced the developmental activation by endogenous heat-shock transcription factors (HSFs) in seeds. Sunflower, and perhaps other plants such as tobacco, differs from the vertebrate animal systems in having at least one specialized HSF with expression and (or) activation patterns strictly restricted to embryos. Our results strongly indicate that HaHSFA9 is a transcription factor critically involved in the developmental activation of Hahsp17.6G1 and in that of similar target genes as Hahsp17.7G4.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In eukaryotes, the heat-shock response and some developmental processes are under the control of a family of conserved DNA-binding proteins known as the heat-shock transcription factors (HSFs).¹ Although in some systems, as in Drosophila melanogaster, this regulation involves a single HSF (1), multigenic families of HSFs participate in vertebrate and in plant systems. These families have different sizes, which, together with particular gene expression and activation patterns for the HSFs, might have consequences in the degree of overlapping of regulatory functions mediated by these factors. The specific role of the different HSFs is mostly unknown, particularly for the plant HSFs, and for involvement in developmental processes, as the regulation of gene expression during embryogenesis (See for example, Ref. 2 and the reviews in Refs. 3 and 4).

In vertebrate systems, three different HSFs (HSF1, HSF2, and HSF3) have ubiquitous expression patterns (for example, Refs. 5 and 6 and the review in Ref. 3). A fourth HSF found in humans displays tissue-specific expression patterns, which suggested specialized functions but not related to embryogenesis (HSF4, Ref. 7). Plants contain the highest number of HSF genes in eukaryotes. This is inferred from in silico analyses from the fully sequenced Arabidopsis thaliana model and from functional analyses of different cloned HSFs in tomato, Arabidopsis, and other plants (reviewed in Refs. 4 and 8 and references therein). Plant HSFs share unique structural and phylogenetic relationships compared with the vertebrate HSFs (9). Fifteen of the 21 putative HSFs from A. thaliana thus contain an insertion of 21 amino acid residues in the oligomerization domain (characteristic of the plant class A HSFs), whereas class B HSFs have no such insertion. Gene expression studies for plant HSFs are very scarce, with only fragmentary data at the mRNA level and even scarcer reports for protein accumulation (also reviewed in Refs. 4 and 8 and references therein). The potential higher specialization of plant HSFs, compared with vertebrate animal systems, is thus mostly unexplored.

Observations from our laboratory indicated the involvement of peculiar HSF(s) and HSE(s) in developmental regulation and of small heat-shock protein (sHSP) gene expression during zygotic embryogenesis in sunflower and other plant systems (10-12). Here we exploited these observations to clone by one-hybrid interaction in yeast HaHSFA9, a class A HSF with a unique embryo-specific gene expression pattern. We show that HaHSFA9 trans-activates two sHSP gene promoters (Hahsp17.7G4 and Hahsp17.6G1), the latter of which is embryo-specific, and that transcriptional regulation by HaHSFA9 in sunflower embryos depends on particular characteristics of the DNA sequences used for cloning (imperfect consensus HSE sequences). We extend similar observations to other endogenous HSFs in transgenic tobacco; thus, we demonstrate the involvement of at least a specialized HSF for sHSP gene regulation during sunflower (and perhaps other plant) embryogenesis.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Construction of the Sunflower Embryo cDNA Library-- mRNA from 14 dpa sunflower embryos (Helianthus annuus, cv. P113HS, seeds from ARLESA Semillas S.A., Spain) was isolated from total RNA using the Poly(A) Quik^R mRNA Isolation kit (Stratagene). We used 5 µg of poly(A) for cDNA synthesis performed with the ZAP-cDNA^R Kit (Stratagene). The cDNA, which included 5' EcoRI adapters, was digested with XhoI. This generated the 3'-cloning site. The digested cDNA was size-fractionated with SizeSep^TM 400-Sepharose CL-4B spun columns (Amersham Biosciences), and directionally inserted within the EcoRI and SalI sites in the polylinker of the pGAD424 vector (Clontech). The primary cDNA library contained 830,000 individual transformants, all with a cDNA insert of average size of 1.3 kb.

One-hybrid Cloning in Yeast-- General methods for one-hybrid cloning and related experimental manipulations in yeast were as described (13). We constructed a yeast strain derivative of YM4271 (Clontech). In this strain, we integrated a HIS3 reporter gene construct ((G4HSE I)x3::HIS3) containing a trimer of the proximal HSE sequences from Ha hsp17.7 G4 (from positions 89 to 57, see Ref. 11). This construct was obtained from plasmid pHISi (Clontech) by insertion, between the XbaI (end-filled with Klenow DNA polymerase) and EcoRI sites, of the DNA fragment generated by annealing of the following two complementary oligonucleotides: 5'-aattcTTCTTCAAGCTTCAAGACAATCCTAGAAATTACTTCTTCAAGCTTCAAGACAATCCTAGAAATTACTTCTTCAAGCTTCAAGACAATCCTAGAAATTAC-3' (top strand) and 5'-GTAATTTCTAGGATTGTCTTGAAGCTTGAAGAAGTAATTTCTAGGATTGTCTTGAAGCTTGAAGAAGTAATTTCTAGGATTGTCTTGAAGCTTGAAGAAg-3' (bottom strand).

For one-hybrid screening, the (G4HSE I)x3::HIS3 reporter yeast strain was transformed with DNA prepared from the embryo cDNA library, after amplification of 1,660,000 primary clones. Five million yeast transformants were plated on SD-His-Leu + 15 mM 3-aminotriazole. After 4-8 days of growth at 30 °C, 24 putative positive yeast clones were selected for further analyses. Four cDNAs encoded the same HSF and are described in this work. The cDNA for one of these clones, HSFA9-36, was subcloned in pBluescript SK+ (Stratagene), thus obtaining pSKHSFA9-36.

RNA Ligase-mediated Amplification of the 5'-cDNA End of HaHSFA9 and Assembly of the Full-length Clone-- The full-length 5'-end of HaHSFA9 cDNA was obtained by rapid amplification mediated by selective ligation to decapped mRNA of an RNA oligonucleotide using T4 RNA ligase (RLM-RACE). We employed the general materials and conditions included in the GeneRacer^TM Kit (Invitrogen). Specific conditions in our case included the use of 1 µg of total RNA from 14 dpa sunflower embryos as starting material. Reverse transcription was performed at 42 °C using random primers. The nested amplification of the full-length cDNA 5'-end was performed at an annealing temperature of 60 °C. We used the following oligonucleotide primer pairs: GeneRacer^TM 5'-primer (for the 1st PCR) or GeneRacer^TM 5'-nested primer (for the 2nd PCR) and HSFA9-RACE primer, which spans between positions 265 and 241 in the noncoding strand of HaHSFA9 cDNA. The PCR-amplified cDNA sequences were cloned in vector pCR^R4-TOPO^R (plasmid pCR^R4-TOPO^R::HSFA9-5'), after addition of 3'-A overhangs under the conditions of the TOPO TA Cloning^R kit (Invitrogen). The full-length HaHSFA9 cDNA clone (plasmid pSKHSFA9-F) was assembled in vector pBluescript SK+ by ligation of a 120-bp DNA fragment from pCR^R4-TOPO^R::HSFA9-5' with the rest of the cDNA sequences in plasmid pSKHSFA9-36. The 120-bp fragment was obtained by PCR amplification (using the primers TOPO-1, 5'-atcATATTCTCCTTCAAAAA-3', and HSFA9-RACE), followed by digestion with StyI (the joining site, at position 122 in the HaHSFA9 cDNA). The EcoRV half-site (lowercase nucleotides) in primer TOPO-1 provided the 5'-insertion point in the vector polylinker sequence.

Ribonuclease Protection Assays-- In vitro transcription, probe purification, and ribonuclease protection assays were performed as described (10). To obtain the HaHSFA9 riboprobe, we cloned (in pBluescript SK+) the EcoRI fragment from plasmid pSKHSFA9-36, and generated pSKHSFA9-RI. The orientation of the insert in pSKHSFA9-RI was such that after linearization with XhoI, transcription from the T3 promoter originated the 507-nucleotide antisense probe used in our experiments.

Antibody Production and Purification for Western Immunodetection-- The antigen was expressed in BL21 Escherichia coli cells from plasmid pRSET-HSFA9BglII. This plasmid included the HaHSFA9 cDNA sequences, between BglII (position 683) and PstI (position 1247), inserted in the vector pRSET-A (Invitrogen). For rabbit immunizations, we purified the antigen by TALON^R metal affinity chromatography (Clontech). The whole serum was affinity-purified using the antigen bound to Immobilon^TM-P (Millipore) (14). The purified antibodies were used at final dilution of 1:4000. For hybridization with the Rabbit anti-HSC70 antibodies (diluted to 1:2000, StressGen Biotechnologies Corp.), blots were stripped using the Re-Blot Plus recycling kit (Chemicon International).

Functional Assays in Yeast and Transient Expression Analyses in Sunflower Embryos-- The expression vector used for the functional assays in yeast contained the full-length HaHSFA9 cDNA, which was excised from plasmid pSKHSFA9-F as a SalI-SacI DNA fragment. This fragment was used to replace the LpHSFA2 SalI-SacI DNA sequences in a plasmid derived from pAD5 as described previously (15). The RSY4 yeast (Saccharomyces cerevisiae) strain and the protocols for galactosidase activity assays were also described (15). To substitute ScHSF1 for HaHSFA9 in RSY4, the HSF1 mutant strain with the ScHSF1 plasmid was transformed with the pAD5-HSFA9 plasmid. The ScHSF1 plasmid was then eliminated by growth in SD-Leu medium followed by plating at low density in SD + 5-fluoro-orotic acid. The final HSF plasmid contents were verified by PCR.

We described previously (16) the conditions used for transient expression in sunflower embryos and the employed GUS reporter plasmids. Double digestion of the full-length cDNA plasmid (pSKHSFA9-F) with EcoRV and SacI allowed the substitution of the GUS sequences in plasmid pBI221 for those of HaHSFA9. In that way we obtained the effector plasmid (p35S::HSFA9) used in the transient activation assays.

Chimeric Gene Expression in Tobacco Transgenic Plants-- By cloning in pBIN19 the SalI-SacI insert of a pBluescript SK-derived plasmid described previously, (1132::GUS(mut P), see Ref. 16), we constructed the binary version of the P mutant chimeric gene. We then generated different primary (T0) tobacco transgenic plant using published procedures (17). The progeny (T1) of plants with the P mutant chimeric gene was also obtained. In parallel, we obtained transgenic plants (T0 and T1) for the previously described binary plasmid with wild type chimeric gene (1132::GUS (WT), see Ref. 17). We compared the developmental (using the T0 plants) and the heat stress-induced expression patterns (using the T1 plants) of different independent transformants (between 4 and 10 per construct) with a similar number (1-3) of integration events. The fluorometric GUS assays during seed maturation and the statistical analyses were as reported previously (11). Heat stress treatments were for 2.5 h at 42 °C followed by recovery for 3 h at 25 °C prior to GUS activity determinations.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Yeast One-hybrid Cloning and Sequence of HaHSFA9-- To isolate trans-acting factors involved in the developmental activation of sHSP gene promoters during the embryogenesis of sunflower, we used the yeast one-hybrid cloning approach (18). Our previous characterization of two such promoters, Hahsp17.6G1 and Hahsp17.7G4, pointed to a similar and peculiar sequence arrangement in their functionally defined HSEs (10, 11). In the case of the Hahsp17.7G4 promoter, we found that the proximal HSE I region would contain overlapping binding sites for HSFs and distinct factors involved in activation during earlier stages of embryogenesis (11). Thus, we decided to use that region as bait that could allow cloning several factors at a time, including the HSF(s) involved in the late developmental activation of both promoters. We directly trimerized these sequences without adding nucleotides that would extend the arrangement of GAA-like and TTC-like HSE core elements with the natural two-nucleotide separation, nnGAAnnTTCnn, etc. We obtained 24 positive colonies. These clones were found to represent different groups. The nucleotide and deduced amino acid sequences showed that one group, with four independent cDNA isolates, encoded a putative transcription factor belonging to the HSF family. Interestingly, in none of them was the GAL4 activation domain cloned in-frame with the deduced amino acid sequences. This indicated that the HSF-encoded sequences were able to activate transcription by themselves, at least in the promoter context of the yeast reporter strain containing the multimerized bait.

We assembled a complete cDNA from the nucleotide sequences determined from the four independent HSF clones (Fig. 1). The four clones included identical nucleotide sequences predicted to encode a protein with DBD and oligomerization domains (with HR-A/B) that identify HSFs (see for example the reviews in Refs. 3 and 19). On the other hand, the nucleotide sequences of these clones maintained the same reading frame to the cloned 5'-ends. Thus, we could not exclude that the predicted HSF protein was N-terminally truncated. Full-length 5'-end cDNA sequences were cloned by RLM-RACE. This added 40 nucleotides that maintained the reading frame without including supplementary initiation codons (Fig. 1).

View larger version (76K): [in this window] [in a new window]   Fig. 1.   Full-length mRNA sequence and characteristics of the predicted HaHSFA9 protein. We assembled the nucleotide sequences of two cDNA clones (HaHSFA9-36 and HFSFA9-38). These cDNAs showed incomplete 5'-untranslated regions with ends indicated by the vertical lines in the 1st (HFSFA9-38) and 2nd sequence lines (HaHSFA9-36). HaHFSFA9-38 and HaHSFA9-36 contained poly(A) tails of 24 and 53 nucleotides, respectively, inserted after the vertical mark in the last sequence line or after the last nucleotide in this figure. The arrow indicates the expected position for the conserved intron present in the genomic sequences. Below the nucleotide sequence, we show the conceptual translation of the HaHSFA9 protein. The stop codon is in boldface. We also indicate putative domains identified by sequence comparison with other HSFs, the DBD including the exceptional amino acid substitution (R in a circle). The oligomerization domain, with overlapping heptad repeats (HR-A/B) indicated by small circles and asterisks; a putative bipartite nuclear localization sequence (NLS); the HR-C repeat, the 21-amino acid insertion between HR-A and HR-B (boldface); and finally the putative AHA motifs (underlined) containing tryptophan residues (W1-W3).

We named the predicted protein HaHSFA9 because its amino acid sequence clearly showed the characteristics of a class A plant HSF (Fig. 1) and because it was most similar to the AtHSFA9 described previously (4). Thus, the DBD and HR-A/B region of HaHSFA9 was 54.5% identical to AtHSFA9 (data not shown). HaHSFA9 showed the distinctive insertion of 21 amino acid residues between the repeat regions HR-A/B in the oligomerization domain (Fig. 1) (9). HaHSFA9 also had a nuclear localization signal located adjacent to the HR-A/B region as well as a C-terminal HR-C region and AHA motifs, which all are classification criteria for class A HSFs (9). However, HaHSFA9 showed some peculiar characteristics compared with other HSFs. This included the very unusual presence in the DBD of an arginine residue in an invariant position encoding glycine in plant and non-plant HSFs (see Arg at position 131 in Fig. 1). Another unusual amino acid change in HaHSFA9 replaced a tyrosine residue conserved in the majority of class A plant HSFs (4), including AtHSFA9 (see Arg at position 114 in Fig. 1). We indicate other unusual structural features (see Fig. 1) as follows: an acidic region N-terminal of the DBD domain (amino acid positions 7-66), and peculiar and putative AHA motifs. These AHA motifs would be related to those found in the activation domains of other class A HSFs (reviewed in Ref. 4). The total number of AHAs and their molecular position (in particular that of C-terminal motif) is different for HaHSFA9.

Embryo-specific Expression of HaHSFA9-- We next investigated the expression patterns of HaHSFA9. We started by determining HaHSFA9 mRNA accumulation patterns using an RNase A protection approach. We performed hybridizations using a collection of total RNA samples from sunflower embryos and from different organs from control and stressed plants. We previously used the same RNA samples for other analyses of developmental mRNA accumulation and to analyze the heat- and water-stress response of other sunflower genes by RNase A protection. This provided positive controls for the different stress treatments and for normal developmental expression (14, 17, 20). For the hybridizations analyzed here, we employed a riboprobe with sequences from the non-coding strand of HaHSFA9. We expected that, after digestion with RNase A, full protection of this probe would produce a protected RNA fragment of 395 nucleotides. Results presented at the top of Fig. 2 show that HaHSFA9 mRNAs accumulated from early zygotic embryogenesis in developing seeds and disappeared very early during germination after seed imbibition. The mRNAs were detected at 8 dpa, and their accumulation slightly increased up to 18 dpa and then started to decrease. These transcripts were barely detectable at 5 dpi and were undetectable at control temperatures in older seedlings (14 dpi) and in different organs of adult plants, as in leaves or stems (Fig. 2, bottom).

View larger version (36K): [in this window] [in a new window]   Fig. 2.   Accumulation patterns of HaHSFA9 mRNAs. Top panel, expression of mRNAs during zygotic embryogenesis and disappearance upon seed germination. The developmental stage is indicated as days post-anthesis (dpa) or days post-imbibition (dpi). Bottom panel, absence of mRNAs in control (C) and heat-stressed (HS) vegetative organs. A marginal accumulation of mRNAs in response to drought stress (DS) could be detected only after a much longer (72 h) autoradiography time than for the embryo samples (18 h). Molecular weight size markers (MW), HpaII-digested, pBluescript (SK+)-labeled DNA. The quality and quantity of total RNA samples used in the ribonuclease A protection assays were verified by electrophoresis in 1% agarose (gel pictures shown below each panel). The scheme at the bottom depicts the 5'-end of transcribed HaHSFA9 sequences (solid line, untranslated region; box, coding region with the predicted conserved intron position as an inverted dark triangle); and the undigested antisense riboprobe (Probe) used in these assays (dashed line). Thick lines below indicate our interpretation of the protected mRNA fragments 1 and 2 (arrowheads in both panels). The question mark indicates a possible alternate mRNA initiation site.

Because heat stress treatments have been shown to induce the mRNA accumulation of some plant HSFs (reviewed in Ref. 4), we investigated such behavior for HaHSFA9. In addition, we investigated the effect of water stress treatment, as the expression of possible target genes for HaHSFA9 is induced during the desiccation stages of embryogenesis (10, 17). Neither stress treatment, in particular heat, induced HaHSFA9 mRNA accumulation after embryogenesis. We only detected very low levels of water stress-induced transcripts in stems and leaves, for which we needed to substantially increase autoradiography times (Fig. 2, bottom).

In these experiments we also detected at least two additional RNA-protected fragments (Fig. 2, arrowhead 2). The most likely explanation for such fragments would be either the detection of unspliced mRNA or that of mRNAs that are shorter from the 5'-end (see the diagram at the bottom of Fig. 2). The presence of an intron in the DBD region of the HaHSFA9 gene is supported by PCR amplification using primers from the cDNA flanking the conserved intron position (Ref. 4 and data not shown).

We obtained specific antibodies against the HaHSFA9 protein (see "Experimental Procedures") and analyzed expression at the protein level during zygotic embryogenesis and in control and stressed plants. The results of Western immunodetection experiments shown in Fig. 3 confirmed most of the results observed at the mRNA level (Fig. 2); the detection of the HaHSFA9 protein in embryos from 8 dpa, and its absence from leaves and stems of control and heat-stressed plants. The protein persisted at low levels in germinated seeds but as observed for the mRNAs only to 5 dpi. We also noticed some differences, as we could not detect HaHSFA9 protein accumulation in response to water stress in vegetative tissues. The detected protein has an apparent molecular mass of 53.7 kDa.

View larger version (54K): [in this window] [in a new window]   Fig. 3.   Seed-specific expression of the HaHSFA9 protein. Western immunodetection confirmed the expression of HaHSFA9 during embryogenesis (from 8 dpa) and mature seeds (22 dpa, top panel, and data not shown). After seed germination, we detected very low levels of the HaHSFA9 protein (arrowheads) only until 5 dpi under control (C) conditions. Heat (H) or drought (D) stress treatments did not re-induce accumulation of the HaHSFA9 protein in vegetative organs, even forcing film exposure to 30 min. In this case as a positive control, we included the 12-dpa embryo sample (bottom panel, E). Molecular mass markers (left side) are in kDa. Sample quality and quantity were confirmed by Ponceau S staining of total proteins (top, P), or by immunodetection using antibodies against HSC70 (bottom, 70).

Functional Analysis of HaHSFA9 in Yeast and in Plant Embryos-- We confirmed the integrity and transcriptional activity of the HaHSFA9 protein(s) encoded by the nucleotide sequences assembled in the full-length cDNA. First, we constructed an appropriate yeast expression plasmid containing the complete coding sequence from the putative first methionine to position 1252 in the 3'-untranslated region. This plasmid was used for functional replacement analysis of ScHSF1, the sole HSF in the yeast S. cerevisiae. The results in Fig. 4A (top) show that this strain was as viable as the original ScHSF1 strain at 28 °C and even at 35 °C. In contrast only ScHSF1 allowed growth at 37 °C. The two strains grew at 28 °C in liquid YPD medium with a similar doubling time (2-2.5 h, data not shown). These results demonstrated functional integrity of the encoded HaHSFA9 protein(s). This is comparable with results described for other class A HSFs from Lycopersicon peruvianum (15).

View larger version (32K): [in this window] [in a new window]   Fig. 4.   The HaHSFA9 cDNA encodes a functional transcription factor. A, top, functional replacement of the yeast ScHSF1 by HaHSFA9. Aliquots (3 µl) of the indicated RSY4 yeast strains were spotted on YPD, starting from mid-log phase cultures, at decreasing cell density (1:10 dilution steps, from left to right). Plates were incubated at the temperatures shown on the left and photographed after 4 days of growth. Bottom, transcriptional activation assays, in yeast, using -galactosidase reporter plasmids containing a minimal promoter with the natural (WT), or mutant (m), HSEs from the Hahsp17.7G4 (G4HSE) and Hahsp17.6G1 (G1HSE) genes. We used the HaHSFA9 yeast strain transformed with the indicated reporter plasmid. -Galactosidase assays were performed, and the values (in duplicate) for three independent transformants per plasmid combination were averaged. The S.E. are represented with bars. B, trans-activation of the Hahsp17.6G1 promoter by the HaHSFA9 effector plasmid in bombarded sunflower embryos. Bars represent mean -glucuronidase (GUS) activity, normalized to luciferase activity (LUC). Reporter plasmids (21) used with (+A9) and without (A9) the HaHSFA9 effector plasmid are indicated. The means ± S.E. are indicated (n  25).

We also used the HaHSFA9 yeast strain to verify the transcriptional activation potential of HaHSFA9 and its dependence on HSE sequence complexity. We constructed lacZ reporter gene plasmids containing the natural HSE region of Hahsp17.7G4 (WT), or a mutant version (m) of it with severe effects on chimeric HSF-dependent gene expression in transgenic plants (m = mutE in Ref. 11). These reporter plasmids contained two HSE repeats (promoter proximal and distal, with its natural spacing in between) in front of minimal yeast promoter, instead the oligomerized proximal HSE repeat used for one-hybrid cloning. HaHSFA9 strains transformed with the WT reporter showed high levels of -galactosidase activity, and this transcriptional activation by HaHSFA9 was severely reduced in the mutant reporter strain (Fig. 4A, bottom, G4HSE). Interestingly the activity measured in this strain was still higher than background values reported for other plant HSFs and mutant HSE reporter plasmids in similar experiments performed without temperature stress (approximately. 1, see Ref. 21). The m HSE array still has a small number of gapped, GAA-, and TTC-like core elements (11) (Fig. 5A). As a consequence, our results could reflect a singular capacity, of HaHSFA9, for binding or/(and) activation from these HSEs. We tested and confirmed this hypothesis by observing transcriptional activation of another lacZ reporter plasmid containing the HSE region from Hahsp17.6G1 (Fig. 4A, bottom, G1HSE). This HSE contains only a promoter-distal, shorter, and gapped array of core elements (10). In this case, the HSE mutations reduced transcriptional activation by HsHSFA9 further than in the case of the G4HSE mutant plasmid and to values more similar than those reported previously for background levels.

View larger version (27K): [in this window] [in a new window]   Fig. 5.   Transcriptional activation in plants of chimeric genes with the Hahsp17.7G4 promoter: embryo-specific effect of HSE mutant P. The scheme on top (A) depicts the promoter context of all chimeric genes used. These were with the natural HSEs WT (1132::GUS, Ref. 16), the previously described "quasi-null" (E, mutE), and perfect (P, mutP) HSE mutants (Refs. 10 and 15, respectively). Core HSE repeats are underlined, with nucleotide substitutions in lowercase. Dots and ovals indicate the crucial nucleotides and non-consensus gaps, respectively, present in the WT HSE and discussed in the text. B, trans-activation by the HaHSFA9 effector plasmid in bombarded sunflower embryos. Bars represent mean -glucuronidase (GUS) activity, normalized to luciferase activity (LUC). The reporter plasmid used with (+A9) and without (A9) HaHSFA9 effector plasmid is indicated. Standard errors and size of samples are as in Fig. 4B. The reduced trans-activation observed with the P mutant plasmid is highlighted with a darker gray shade. Bottom, chimeric gene expression in transgenic tobacco. C, (seedlings): lack of effect of mutant P in control (C) and heat stress (HS) conditions. D, embryogenesis, reduced activity of the P mutant gene at 28 dpa. In this case, the statistical significance for the effect is indicated with the same shade as in B.

These results suggested that HaHSFA9 might activate transcription from the Hahsp17.6G1 promoter in plant seeds. We performed transient expression assays in sunflower embryos. We used two chimeric genes that were previously analyzed in transgenic plants: 1486::GUS and 1486(m)::GUS. Such genes only differ in three crucial nucleotide substitutions, located in the HSE region of Hahsp17.6G1. These substitutions impaired transcriptional activation during desiccation stages of zygotic embryogenesis (20). The results shown in Fig. 4B demonstrated that HaHSFA9, expressed from an appropriate plant effector plasmid, efficiently trans-activated the 1486::GUS gene. Mutations in the 1486(m)::GUS gene abolished transcriptional activation by HaHSFA9.

Embryo-specific Effect of Mutations That Improve the HSE in the Hahsp17.7G4 Promoter-- In sunflower, we described previously that there is another sHSP gene, Hahsp18.6G2, that is not transcriptionally activated in developing embryos, but it is responsive to heat stress. The HSE arrays of Hahsp18.6G2 are more "perfect" than those in Hahsp17.6G1 or Hahsp17.7G4, in the sense that they lack gaps between the core repeats, which are thus more similar to the aGAA/TTCt consensus sequences (10). These observations prompted us to study the effect on transcriptional activation by HaHSFA9 and related factors of nucleotide substitutions that make HSEs more similar to those in Hahsp18.6G2.

We analyzed these mutations in the natural context of the Hahsp17.7G4 promoter because its proximal HSE sequences were used to clone HaHSFA9, and Hahsp17.7G4 is both developmentally activated and heat stress-responsive. In Fig. 5A, we depict the perfect (P) Hahsp17.7G4 HSE mutant compared with the unaltered (WT) and strong negative mutant (m) HSE versions. We first investigated the effect of mutant HSEs on transient activation by HaHSFA9 in sunflower embryos (see Fig. 5B). We confirmed that HaHSFA9 very efficiently activated transcription of the WT gene. In these assays, the mutant HSE completely abolished transcriptional activation by HaHSFA9, as the reporter activity observed with the HaHSFA9 effector plasmid did not differ from basal control levels (F = 0.262, p = 0.61). The P mutant HSE produced two contrasting effects; it significantly increased basal activity levels (F = 40.01, p = 0.0001) but also clearly reduced transcriptional activation by HaHSFA9 (F = 9.82, p = 0.002). Thus, HaHSFA9 differs from other HSFs present in sunflower embryos in its capacity of binding and (or) transcriptional activation from the P HSE.

We finally investigated the effect of the P mutant HSE on the developmental activation and heat stress response of the chimeric gene in tobacco plants, where we previously showed a faithful reproduction of transcriptional regulation of Hahsp17.7G4 (11, 17). The results in Fig. 5C show that, compared with the WT gene, mutant P was not affected in the heat stress response as determined by experiments performed with seedlings. Statistical analyses determined that both the control (F = 0.06, p = 0.81) and induced (F = 0.002, p = 0.96) activity levels of the two genes were similar. In contrast, the WT and P genes differed in their developmental activation during zygotic embryogenesis (Fig. 5D). The P gene showed reduced activation during desiccation stages of late embryogenesis (28 dpa, F = 36.5, p = 0.0001) but not earlier (16 dpa, F = 1.67, p = 0.2). We conclude that the effect of mutant P is not only embryo-specific but also stage-specific. Furthermore, such decrease is analogous to what was observed in sunflower embryos for transcriptional activation of the same genes by HaHSFA9 (compare Fig. 5, B and D). The transgenic plant system (Fig. 5D) showed a higher reduction in the activity of the P gene, most likely explained by differences between transient (Fig. 5B) and stable (Fig. 5D) expression conditions.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Embryo Specificity of HaHSFA9-- The embryo-specific expression pattern observed for HaHSFA9 is supported by the results of mRNA (Fig. 2) and protein (Fig. 3) accumulation assays. RNase A protection (Fig. 2) combines sensitivity comparable with reverse transcriptase-PCR with the highest specificity (22). In addition to the data presented in Figs. 2 and 3, we have confirmed the results for mRNA using reverse transcriptase-PCR, and we were unable to detect the HaHSFA9 protein in roots from adult plants (data not shown). We thus explored most vegetative organs and tissues in young and adult plants, in the absence and presence of stress (heat or drought). We conclude that, within the detection limits in our experimental conditions, the HaHSFA9 protein is expressed only in embryos, and it disappears shortly after seed germination. The accumulation of the HaHSFA9 protein precedes and appears connected to the embryo desiccation phase, when potential target genes reach maximal expression levels (see for example Ref. 10 and references therein).

The strict embryo specificity of protein accumulation (Fig. 3) restricts the function(s) of HaHSFA9, which appears as a regulator of developmental gene expression during zygotic embryogenesis in sunflower. This would be different from animal systems. In humans, only HSF4 has strict tissue-specific protein expression patterns in the lung and brain. However, the functions proposed for hHSF4 in these organs would not be related with embryogenesis (7). In vertebrates protein and mRNA accumulation, assays showed that HSF1, HSF2, and HSF3 have ubiquitous expression patterns (reviewed in Ref. 3). Even in the case of HSF2, for which functions related with embryo development have been proposed (2), expression is ubiquitous at low levels (5). This suggests additional non-embryonic functions of HSF2 in most cell types. HSF1 and HSF3 also have other functions in addition to their involvement in heat-shock tolerance. Specific interactions, with different transcription factors, could confer distinct functional specificity to the ubiquitous expression pattern of animal HSFs. An example of this is the interaction between c-Myb factor and HSF3 (3, 23).

Expression analyses for the more numerous plant HSFs are still scarce. For example, for the 21 HSFs predicted from the genomic sequence of A. thaliana there are only fragmentary RNA studies (some of them limited to the description of different ESTs). Protein accumulation analyses using specific antibodies have been reported only for three HSFs from tomato: LpHSFA1, LpHSFA2, and LpHSFA3. Some peculiarities have been noted from the published information (recently reviewed in Refs. 4 and 8 and see references therein) on plant HSFs as follows: 1) the existence not only of ubiquitous (as LpHSFA1) but also of heat-inducible HSFs such as LpHSFA2; 2) the detection of LpHSFA3 in proliferating cell cultures (21) but not in unstressed leaves (Ref. 4, see the unpublished observations). Thus, the higher number of plant HSFs could potentially increase the complexity of interactions with other transcription factors and provide additional functional diversification by evolving distinct specific expression patterns (as demonstrated in this work for HaHSFA9). Previous evidence supporting functional specialization of plant HSFs is scarce and connected only to thermotolerance (24, 25) or to cell death defense (26). This was deduced from the specific effects observed by gain of function (24) or, more recently, by loss of function of different HSFs (25, 26).

Functional Characteristics and Validation of HaHSFA9-- Perhaps the most unusual characteristic of transcriptional activation by HaHSFA9 is the detrimental effect of HSE mutations that improve the consensus sequences, by replacing the natural non-consensus nucleotides ("gaps") located between the GAA repeats in the HSEs of Hahsp17.7G4 (Fig. 5). This observation is unprecedented for an HSF. Moreover, as in transgenic tobacco during late embryogenesis the mutant P gene also showed reduced activity, compared with WT, we infer that endogenous HSFs, with similar transcriptional activation properties as HaHSFA9, are present in this system. Furthermore, the seed specificity of the effect of mutant P in tobacco would be consistent with HSF(s) with functions mostly embryonic, as proposed above for HaHSFA9 based in its peculiar expression pattern.

The increased basal expression levels observed for the P mutant gene in sunflower embryos, compared with that of WT and m versions (Fig. 5B), would indicate that additional HSFs are expressed in these embryos. This is consistent with reports (reviewed in Ref. 4) showing that cDNAs and ESTs for different plant HSFs can be cloned from samples containing embryos² (seeds ± seedpods). Contrary to HaHSFA9, these additional HSFs would trans-activate the P mutant gene more efficiently than the WT gene (Fig. 5B). As another counterexample to HaHSFA9, we had shown that the same WT and P genes were equally trans-activated by LpHSFA1, in similar experiments performed in sunflower embryos (16). Thus, the DNA binding (and/or) activation characteristics of HaHSFA9 would be exceptional among plant HSFs.

Previous work from our laboratory suggested the critical involvement of atypical HSFs in the seed-specific expression of Hahsp17.6G1 (10, 12, 20) and in the developmental regulation of Hahsp17.7G4 (11). The gapped HSEs in both promoters were necessary for their transcriptional activation during late embryogenesis in transgenic tobacco (11, 20). In the case of Hahsp17.7G4, such involvement was inferred from the different effect of very specific HSE mutations on the heat response and on the developmental regulation (11). On the other hand, the Hahsp17.6G1 promoter does not respond to heat shock (10, 20), and its HSE is very selective for activation by the heterologous HSFs LpHSFA1 and LpHSHA2 (12). These HSFs, when individually tested with the Hahsp17.6G1 HSE, activated transcription only marginally either in yeast cells or in sunflower embryos (12). This contrasts the efficient trans-activation observed with HaHSFA9 in the analogous experiments shown in Fig. 4, A (bottom) and B, respectively. LpHSFA1 or LpHSHA2 also efficiently trans-activated through different HSEs using the same yeast (21) and sunflower systems (16). Our previous studies (10, 11, 12, 16, and 20) did not anticipate the following two properties of HSFs involved in developmental regulation in plants: 1) the embryo specificity, and 2) the detrimental transcriptional effect of HSE consensus improvement. These two unique properties, together with the efficient activation through the Hahsp17.6G1 HSE, have been crucial for functional identification of HaHSFA9 in sunflower and to confirm the occurrence of equivalent HSFs in tobacco. Thus, HaHSFA9 is most likely involved in the developmental regulation of a peculiar subset of sHSP genes, including Hahsp17.6G1 and Hahsp17.7G4, during zygotic embryogenesis. The findings reported here open the possibility of investigating functions that have been proposed for such sHSP genes (see for example, Ref. 17). This might be achieved by gain-of-function or loss-of-function approaches in transgenic plants using HaHSFA9.

	ACKNOWLEDGEMENT

We thank Pilar Bazaga for excellent technical work.

	FOOTNOTES

^* This work was supported in part by Grant BIO99-794 (to J. J.) from the Spanish Ministerio de Ciencia y Tecnología, Plan Nacional de I+D+I.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence(s) reported in this paper has been submitted to the GenBank^TM/EBI Data Bank with accession number(s) AY099451.

Present address: Cardiovascular Research Institute, University of California, San Francisco, 505 Parnassus Ave., San Francisco, CA 94143-0130.

^§ Supported by Ph.D. fellowships from the Spanish "Ministerio de Educación y Cultura."

^¶ Present address: Institute of Developmental and Molecular Biology and Dept. of Biology, Texas A & M University, College Station, TX 77843-3155.

To whom correspondence should be addressed. Tel.: 34-954-624711, ext. 145; Fax: 34-954-624002; E-mail: fraga@cica.es.

Published, JBC Papers in Press, September 12, 2002, DOI 10.1074/jbc.M207330200

² C. Almoguera, A. Rojas, J. Díaz-Martín, P. Prieto-Dapena, R. Carranco, and J. Jordano, unpublished observations for sunflower.

	ABBREVIATIONS

The abbreviations used are: HSF, heat-shock factor; sHSP(s), small heat-shock protein(s); HSE, heat-shock element; DBD, DNA-binding domain; HR, hydrophobic heptad repeats; AHA, aromatic, large hydrophobic and acidic amino acid motif; dpi, days post-imbibition; dpa, days post-anthesis; m, mutant; WT, wild type; P, Perfect.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				S. K. Baniwal, K. Y. Chan, K.-D. Scharf, and L. Nover Role of Heat Stress Transcription Factor HsfA5 as Specific Repressor of HsfA4 J. Biol. Chem., February 9, 2007; 282(6): 3605 - 3613. [Abstract] [Full Text] [PDF]

				S. Kotak, E. Vierling, H. Baumlein, and P. v. Koskull-Doring A Novel Transcriptional Cascade Regulating Expression of Heat Stress Proteins during Seed Development of Arabidopsis PLANT CELL, January 1, 2007; 19(1): 182 - 195. [Abstract] [Full Text] [PDF]

				P. Prieto-Dapena, R. Castano, C. Almoguera, and J. Jordano Improved Resistance to Controlled Deterioration in Transgenic Seeds Plant Physiology, November 1, 2006; 142(3): 1102 - 1112. [Abstract] [Full Text] [PDF]

				J. Diaz-Martin, C. Almoguera, P. Prieto-Dapena, J. M. Espinosa, and J. Jordano Functional Interaction between Two Transcription Factors Involved in the Developmental Regulation of a Small Heat Stress Protein Gene Promoter Plant Physiology, November 1, 2005; 139(3): 1483 - 1494. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 277/46/43866    most recent M207330200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Almoguera, C. Articles by Jordano, J. Search for Related Content PubMed PubMed Citation Articles by Almoguera, C. Articles by Jordano, J. Social Bookmarking