TY - JOUR
AU - Parcy, François
AB - Abstract The Arabidopsis thaliana genome contains approximately 80 genes encoding basic leucine zipper transcription factors, divided into 11 groups. Abscisic Acid-Insensitive 5 (ABI5) is one of the 13 members of group A and is involved in ABA signalling during seed maturation, and germination. Seven other members of this group are expressed during seed maturation, but only one of them (Enhanced Em Level, EEL) has been functionally characterized during this developmental phase. Since EEL and two other group A genes, AtbZIP67 and AREB3 (ABA-Responsive Element Binding protein 3), display similar mRNA temporal expression in whole siliques, it is suspected that they might share some overlapping functions. To address this question, the proteins' tissular and subcellular localization in transgenic Arabidopsis were precisely characterized, using translational fusions with a green fluorescent protein (GFP) expressed under the corresponding bZIP promoter. It was found that the three fusion proteins were expressed with a largely overlapping pattern and constitutively localized in the nuclei. An RNA interference approach (RNAi) was then used to knock out the expression of all three genes simultaneously. Endogenous EEL, AREB3, and AtbZIP67 transcripts could be specifically reduced, but no visible defects could be observed during seed maturation. ABI5, bZIP transcription factors, multigene family, RNA interference, seed maturation This paper is available online free of all access charges (see http://jxb.oupjournals.org/open_access.html for further details) Introduction Many transcription factors encoded by the Arabidopsis genome belong to multigene families (Riechmann et al., 2000). Functional overlap or redundancy between members of a given family often complicates functional analysis of individual factors (Pelaz et al., 2000; Stracke et al., 2001; Liscum and Reed, 2002; Messenguy and Dubois, 2003). Since redundancy mostly occurs between simultaneously expressed factors, a good characterization of expression patterns is important to identify possible functional overlap. Once concomitant expression is established, the function of the co-expressed factors can be assayed genetically by using a combination of mutations or multiple gene silencing. This strategy was applied to members of the basic leucine zipper family. Basic leucine zipper transcription factors (bZIP) are characterized by evenly spaced leucine residues allowing dimerization and a DNA binding basic domain (Hurst, 1995). Several bZIP subfamilies have been described in Arabidopsis based on the presence of conserved domains and gene structure (Jakoby et al., 2002). One of these subfamilies (called group A) contains 13 genes including the ABI5, AREB, and ABF genes involved in abscisic acid (ABA) signalling. There was interest in understanding their role during the developmental phase of seed maturation (Goldberg et al., 1994). This phase follows embryo morphogenesis and prepares the seed for desiccation. The embryo enters a quiescent stage, accumulates storage compounds and acquires desiccation tolerance. Among the seven group A bZIP genes expressed during seed maturation, only two have been shown to play a role during this process: ABI5, which participates in ABA signalling and positive regulation of Late Embryogenesis Abundant (LEA) genes (Finkelstein and Lynch, 2000; Lopez-Molina and Chua, 2000; Lopez-Molina et al., 2002; Carles et al., 2002), and EEL which counteracts ABI5 action on some LEA genes to fine-tune their expression (Bensmihen et al., 2002). ABI5 and EEL expressions have been characterized at the mRNA level in whole seeds. ABI5 level peaks towards the end of seed maturation, whereas EEL expression reaches a plateau at mid-maturation and decreases later (Bensmihen et al., 2002). Two other bZIP genes of unknown function, named AREB3 and AtbZIP67, display a temporal expression pattern similar to EEL (Bensmihen et al., 2002). However, the spatial expression and intracellular localization of the corresponding proteins are unknown. To gain insight into these proteins' localization, GFP-tagged versions of EEL, AREB3, and AtbZIP67 proteins were expressed under their own promoter and their expression in the maturing embryo was analysed. Since they were found to display a largely overlapping expression pattern, triple RNAi lines were also generated and a concomitant decrease of expression of the three genes was observed. Materials and methods Plant transformation and growth ProbZIP∷GFP-bZIP and RNAi constructs were introduced into the Columbia 0 (Col0) accession by Agrobacterium transformation using the floral dip method (Clough and Bent, 1998). All plants were grown in greenhouse under long day conditions (22 °C, 16 h light). ProbZIP∷GFP-bZIP transformants were selected by spraying young seedlings with Basta herbicide (glufosinate, 60 mg l−1). RNAi T1 seeds were selected on germination medium (as described in Bensmihen et al., 2002) containing 1% sucrose, 50 μg l−1 kanamycin (Sigma) and 400 mg l−1 of the antibiotics amoxicilline/clavulanic acid (1 g 100 mg−1) (Augmentin, SmithKline Beecham Laboratories). DNA constructs Vectors: ProbZIP∷GFP-bZIP constructs were built in the pSBright vector. This new binary vector was constructed by assembling a SpeI/XbaI/HindIII multiple cloning site upstream of a smRS-GFP gene (Davis and Vierstra, 1998) and a Gateway™ Reading Frame A (RfA) cassette (Invitrogen). ABI5, EEL, AREB3, and AtbZIP67 cDNA (as described in Bensmihen et al., 2002) were first cloned in the pDONR201 vector (Invitrogen), prior to their recombination in pSBright following the manufacturer's instructions. The PSBright map is available online at http://www.isv.cnrs-gif.fr/jg/alligator/othervectors.html. BZIP promoters: 1760 bp upstream of the ABI5 ATG were amplified with the 5′-TTACTAGTAGTTTTTTGGCTATTAGAAACACTTGATA-3′ (ABI5for) and 5′-TTAAAGCTTTACCATTTAACAACTGCATCATATACACAA-3′(ABI5rev) primers from the F2H17 BAC (ABRC, Ohio State University, Columbus) and introduced in pSBright using the SpeI and HindIII polylinker sites. 2080 bp upstream of the EEL ATG were amplified with the 5′-ATTACTAGTCCATGCACGTAATTCCTCTTAAGAT-3′ (EEL for) and 5′-TTTAAGCTTACCCATATATGTAGCCTTTACACAGA-3′ (EELrev) primers from the T3K9 BAC (ABRC, Ohio State University, Columbus) and introduced in pSBright using the SpeI and HindIII polylinker sites. 1830 bp upstream of the AREB3 ATG were amplified with the 5′-TTAACTAGTTCAACGACACAGCTGAAGAATGAT-3′ (AREBfor) and the 5′-TAAAAGCTTATCCATAGGCTTTTGTAGCGGACAA-3′ (AREBrev) primers from the T8M6 BAC (ABRC, Ohio State University, Columbus) and introduced in pSBright using the SpeI and HindIII polylinker sites. 1900 bp upstream of the AtbZIP67 ATG were amplified with the 5′-TTACTAGTAAACTTGAAATTGTCTTGTTTGGCCTCCTA-3′ (bZIP67for) and 5′-AAACTAGTCGACATCGTTTGGTAGACCTATAAATCTTGAA-3′ (bZIP67rev) primers from the F14L2 BAC (ABRC, Ohio State University, Columbus) and introduced in pSBright using the SpeI polylinker site. Each promoter amplification was performed twice independently using the Taq Hifi DNA polymerase (Invitrogen). Both amplification products were subsequently used for in planta transformation and characterization. RNAi construct: 140 bp of EEL, AREB3, and AtbZIP67 specific probes (as described in Bensmihen et al., 2002) were amplified respectively with: 5′-ATTTCTAGATTGTCTCGAGCTACTCCTTCTTAATCT-3′ (EELifor) and 5′-ATAAACCTGCAGCTAATAGAACCCATATATGTAGC-3′ (EELirev); 5′-TTCTCATGCATGAAAAGTCTGTACCTCGCAA-3′ (AREBifor) and 5′-AAAGGATCCAATCGAAGGAAATGGCA-3′ (AREBirev); 5′-TGAAGATCTCTAGCGAACTTGTGGACAACCGTTGAA-3′ (bZIP67ifor) and 5′-AAAATTAAGCTTTCAAGGTACCCCTCGTCGACCGTTTTCT-3′ (bZIP67irev) primers. The three fragments were then digested with PstI (EEL), NsiI, and BamHI (AREB3), and BglII (AtbZIP67) respectively, ligated and re-amplified with EELifor and bZIP67irev, leading to the EEL-AREB3-AtbZIP67 fragment. This fragment was then introduced as an inverted-repeated sequence in pHannibal (Wesley et al., 2001) using the XhoI/KpnI and XbaI/HindIII cloning sites, respectively (EELifor introduces XbaI and XhoI restriction sites, bZIP67irev introduces HindIII and KpnI sites). This ‘RNAi fragment’ was then transferred in pART27 using the NotI cloning site. RNA handling RT-PCR: Total RNA was extracted from seeds using a phenol–chloroform protocol (Parcy et al., 1994). Total RNA was then cleaned using the ‘Clean Up’ protocol from the RNeasy kit (Qiagen). Reverse transcription was performed with the SuperscriptII enzyme (GibcoBRL), following the manufacturer's instructions. One μg of total RNA was then used for the PCR reactions. Primers used were as follow: 5′-AATGGTGTCTCAGTCTTCTTTGATGG-3′ (AREBforRT) and 5′-TTTAGAGATCAGAAAGGAGCCGAG-3′ (AREBrevRT) for AREB3; 5′-TCTGGAAAACCACTAGGAAGCAT-3′ (FP1079) and 5′-AAGAAGAGTCTTTAGGATCAGAGAG–3′ (EELrevRT) for EEL; 5′-TTATAACCCCGAGTTTGGAGTTG–3′ (67forRT) and 5′-AATTCCAACTCCAGTTCCACAG-3′ (67revRT) for AtbZIP67; 5′-GTTTAGAGTGGACAACTCGGGTTCC-3′ (FP1086) and 5′- GGGGAAGGAAAAGAGTAGTGG-3′ (FP1089) for ABI5; 5′-ATGCCCCAGGACATCGTGATTTCAT-3′ (EF1RT1) and 5′-TTGGCGGCACCCTTAGCTGGATCA–3′ (EF1RT2) for EF1α. Each amplification was performed with an annealing temperature of 56 °C and 25 amplification cycles. Microscopy techniques Staging of the individual siliques was done by tagging individual flowers on the day of pollination. Embryos were taken out of the seed coat by gently squashing seeds in a water drop between a glass slide and a cover slip. A TCS SP2 confocal microscope from Leica was used. GFP was excited with the 488 nm line of the Argon laser (34–38% of power at approximately 4/10 gain). Chlorophyll excitation was reinforced using the 633 nm line of a Helium/Neon laser (30% of power). The objective used was a HCX PL Apo CS 40X, NA 1.25. The images were scanned on a 15–30 μm width with 1 μm increment in Z. The images are averaging projections of the Z-series, provided by the Leica LCS software. Results Construction of pSBright vector to express translational GFP fusions in planta A versatile binary vector was built first to allow expression of translational fusion between any protein and a soluble version of GFP (Davis and Vierstra, 1998) under the control of any promoter of interest. For this, a small multiple cloning site, the GFP coding sequence, a Gateway recombination cassette, and a transcriptional terminator was assembled (Fig. 1a). Any promoter can be inserted in the multiple cloning site and cDNAs can be recombined in frame with the GFP sequence. This strategy was applied to the ABI5, EEL, AREB3, and AtbZIP67 genes, thereby generating four constructs designated ProbZIP∷GFP-bZIP. Fig. 1. View largeDownload slide Strategy for expression of GFP-tagged bZIP proteins in planta. (a) Schematic map of pSBright binary vector. The gateway cassette (GTW) enables in-frame recombination of any cDNA with the soluble modified GFP gene (GFP). The multiple cloning site (SpeI, XbaI, and HindIII) allows introduction of the promoter of interest. This vector confers resistance to the Basta herbicide in planta. LB, left border of T-DNA; RB, right border of T-DNA; NOS-T, nopaline synthase terminator. (b) GFP-ABI5 protein location in planta. GFP-ABI5 fluorescence in axis (A) and cotyledon (B) of ProABI5∷GFP-ABI5 15 DAP embryo. Green (500–530 nm) emission (C, E) and green and red (640–685 nm) emissions overlay (D, F) from transgenic (C, D) and wild-type (E, F) cotyledons, showing nuclei (big spots in C, D) and chloroplasts fluorescence. Fluorescence (550–570 nm, coded blue) from di-hydroethidium staining of nuclei from cotyledons tissues (H) coincides with GFP emission (G) as shown by the overlay of both emission channels (light blue in J). Chlorophyll emission is also shown (I). Bar is 40 μm in (A, B, C, D, E, F) and 20 μm in (G, H, I, J). Fig. 1. View largeDownload slide Strategy for expression of GFP-tagged bZIP proteins in planta. (a) Schematic map of pSBright binary vector. The gateway cassette (GTW) enables in-frame recombination of any cDNA with the soluble modified GFP gene (GFP). The multiple cloning site (SpeI, XbaI, and HindIII) allows introduction of the promoter of interest. This vector confers resistance to the Basta herbicide in planta. LB, left border of T-DNA; RB, right border of T-DNA; NOS-T, nopaline synthase terminator. (b) GFP-ABI5 protein location in planta. GFP-ABI5 fluorescence in axis (A) and cotyledon (B) of ProABI5∷GFP-ABI5 15 DAP embryo. Green (500–530 nm) emission (C, E) and green and red (640–685 nm) emissions overlay (D, F) from transgenic (C, D) and wild-type (E, F) cotyledons, showing nuclei (big spots in C, D) and chloroplasts fluorescence. Fluorescence (550–570 nm, coded blue) from di-hydroethidium staining of nuclei from cotyledons tissues (H) coincides with GFP emission (G) as shown by the overlay of both emission channels (light blue in J). Chlorophyll emission is also shown (I). Bar is 40 μm in (A, B, C, D, E, F) and 20 μm in (G, H, I, J). Localization of GFP-tagged bZIP during embryo maturation The pSBright vector, as well as the feasibility of this approach in embryos using a ProABI5∷GFP-ABI5 construct were tested first. The ABI5 expression pattern and subcellular location had been characterized (Brocard et al., 2002; Lopez-Molina et al., 2002) : ABI5 expression peaks at the end of seed maturation and the ABI5 protein appears to be constitutively localized in the nucleus. A construct containing 1760 bp of the ABI5 promoter upstream of the GFP-ABI5 fusion was introduced into wild-type Arabidopsis and its expression was analysed during seed development. In these conditions, seed maturation occurs between 7 d and 15 d after pollination (DAP), 15 DAP corresponding to the dry-seed stage. 14 and 15 DAP embryos from two independent ProABI5∷GFP-ABI5 lines were dissected out of the seed coat. As shown in Fig. 1b, bright green spots were detected in both the axis (Fig. 1b, A) and cotyledons (Fig. 1b, B) from the transgenic embryos. Some smaller spots were also visible. However, they appeared both in transgenic (Fig. 1b, C) and wild-type (Fig. 1b, E) plants. The coincidence of these green spots with chlorophyll emission (Fig. 1b, D, F) suggested they correspond to the chloroplasts' own fluorescence. The brighter and bigger spots are only seen in transgenic plants (Fig. 1b, C) and do not overlap with chlorophyll (Fig. 1b, D). It was concluded that they represent the GFP signal. The form and localization of these spots in the cell strongly suggest they correspond to nuclei. This hypothesis was confirmed by co-localization of this signal (Fig. 1b, G, J) with fluorescence from the DNA-specific dye, di-hydroethidium (Fig. 1b, H). The nuclear localization depends on the ABI5 coding sequence, as GFP alone expressed under the same ABI5 promoter is located both in the nucleus and the cytoplasm (data not shown). These experiments established the functionality of the pSBright vector and demonstrated that it can be used to analyse protein expression in embryos. This vector was then used for similar constructs using the EEL, AREB3, and AtbZIP67 promoters and cDNAs. EEL, AREB3, and AtbZIP67 genes have been characterized previously (Choi et al., 2000; Bensmihen et al., 2002) and are known to be expressed in maturing embryos (Bensmihen et al., 2002). However, no information about the protein localization in the embryo has been presented. The fluorescence of the GFP-bZIP fusions was observed in maturing embryos for all ProbZIP∷GFP-bZIP constructs. None of the three constructs generated fluorescence in the embryos before 7 DAP (data not shown). GFP-AtbZIP67 fluorescence was first detected in the cotyledons and axis of 8 DAP embryos (Fig. 2I), whereas GFP-AREB3 and GFP-EEL fluorescence appeared 1 d later, in the whole embryo for AREB3 (Fig. 2E), but only in the cotyledons for EEL (Fig. 2A). From 10 DAP to 13 DAP, fluorescence was then detected for the three GFP-bZIP throughout the embryo, in both cotyledons and axis (Fig. 2B, C, F, G, J, K). At 13 DAP, GFP-EEL and GFP-AREB3 fluorescence persisted in the whole embryo (data not shown) whereas GFP-AtbZIP67 fluorescence was restricted to cotyledons (Fig. 2L). Finally, at 14 DAP, only GFP-AREB3 fluorescence remained detectable (Fig. 2H), whereas GFP-EEL was no longer visible (Fig. 2D). No GFP-bZIP fluorescence was detected in dry seeds. Throughout their expression window, fluorescence was specifically detected in the nuclei, indicating that all three proteins were constitutively nuclear. Fig. 2. View largeDownload slide GFP-bZIP fluorescence pattern during embryo maturation in ProEEL∷GFP-EEL, ProAREB3∷GFP-AREB3, and ProAtbZIP67∷GFP-AtbZIP67 lines. GFP-bZIP fluorescence was observed using confocal microscopy in both cotyledons (B, D, F, H, J, L) and axis (C, G, K) from isolated embryos at the indicated time of development (DAP: days after pollination). For the earliest stage, part of whole embryos is represented (A, E, I), showing cotyledons (Cy) and axis (Ax). Images are overlays of green (500–530 nm) and red (640–685 nm) emission channels. Four independent transgenic lines were observed for each construct. Bar is 40 μm. Fig. 2. View largeDownload slide GFP-bZIP fluorescence pattern during embryo maturation in ProEEL∷GFP-EEL, ProAREB3∷GFP-AREB3, and ProAtbZIP67∷GFP-AtbZIP67 lines. GFP-bZIP fluorescence was observed using confocal microscopy in both cotyledons (B, D, F, H, J, L) and axis (C, G, K) from isolated embryos at the indicated time of development (DAP: days after pollination). For the earliest stage, part of whole embryos is represented (A, E, I), showing cotyledons (Cy) and axis (Ax). Images are overlays of green (500–530 nm) and red (640–685 nm) emission channels. Four independent transgenic lines were observed for each construct. Bar is 40 μm. These results show that all three proteins are expressed during seed maturation with slightly different but overlapping expression patterns, suggesting they may perform overlapping functions. For this reason, an attempt was made to silence the expression of all three genes simultaneously, using an RNA interference approach. Specific endogenous bZIP gene extinction using a ‘triple’ RNA interference (RNAi) approach An attempt to silence the three EEL, AREB3, and AtbZIP67 genes was made by assembling three 140 bp DNA fragments amplified from specific regions of EEL, AREB3, and AtbZIP67 genes. The resulting concatemer was introduced as an inverted-repeated fragment in the pHannibal vector, as shown on Fig. 3a, to generate an RNAi construct as described by Wesley et al. (2001). Fig. 3. View largeDownload slide RNAi approach to trigger EEL, AREB3 and AtbZIP67 silencing. (a) 140 bp of specific regions from the EEL, AREB3, and AtbZIP67 cDNAs were assembled in a single fragment introduced in the pHannibal vector (Wesley et al., 2001). This enables the expression of an inverted-repeated version of this fragment under the 35S promoter in planta. (b) 11 DAP F1 embryos from crosses between RNAi lines and the ProAREB3 ∷GFP-AREB3 line. Progeny from RNAi lines #17.3 (C) and #10.3 (D) crosses displayed severe GFP reduction in both axis (C, D) and cotyledons (not shown), whereas progeny from #13 line cross still displayed high level of GFP fluorescence (B). Progeny from Col0 cross did not show any significant GFP fluorescence reduction (A). Seven to 10 embryos per cross were observed. Bar is 40 μm. Fig. 3. View largeDownload slide RNAi approach to trigger EEL, AREB3 and AtbZIP67 silencing. (a) 140 bp of specific regions from the EEL, AREB3, and AtbZIP67 cDNAs were assembled in a single fragment introduced in the pHannibal vector (Wesley et al., 2001). This enables the expression of an inverted-repeated version of this fragment under the 35S promoter in planta. (b) 11 DAP F1 embryos from crosses between RNAi lines and the ProAREB3 ∷GFP-AREB3 line. Progeny from RNAi lines #17.3 (C) and #10.3 (D) crosses displayed severe GFP reduction in both axis (C, D) and cotyledons (not shown), whereas progeny from #13 line cross still displayed high level of GFP fluorescence (B). Progeny from Col0 cross did not show any significant GFP fluorescence reduction (A). Seven to 10 embryos per cross were observed. Bar is 40 μm. Twenty-two independent RNAi lines were generated. To select those displaying the most efficient silencing, each of them was crossed to the ProAREB3∷GFP-AREB3 line described above. The RNAi construct is supposed to trigger the degradation of the endogenous transcript as well as any chimeric transcript containing the targeted sequence (Voinnet et al., 1998). As the RNAi effect is dominant, efficient RNAi lines should trigger the degradation of the GFP-bZIP transcript and the loss of GFP fluorescence in the F1 progeny of the cross. This test enabled selection of the most efficient RNAi lines. As shown in Fig. 3b, some RNAi lines, such as #13, led to a poor diminution of GFP fluorescence in the F1 embryo (Fig. 3b, B), whereas others, such as #17.3 (Fig. 3b, C) or #10.3 (Fig. 3b, D) led to a severe reduction of GFP fluorescence. However, this fluorescence loss is not complete and some signal could still be detected in the root tip of the F1 embryo (Fig. 3b, C, D). Crossing of #13, #17.3, and #10.3 lines to ProEEL∷GFP-EEL and ProAtbZIP67∷GFP-AtbZIP67 lines gave similar results (data not shown). This fluorescence loss could not be observed by crossing these lines to the ProABI5∷GFP-ABI5 line (data not shown), showing that the silencing is specific for the three targeted bZIP genes. To test whether the silencing could also be observed at the mRNA level, the expression of EEL, AREB3, and AtbZIP67 was monitored by RT-PCR in the #10.3, #17.3, and #13 T3 homozygous RNAi lines. As shown in Fig. 4, the level of EEL, AREB3, and AtbZIP67 transcripts seemed poorly affected in the siliques of the #13 RNAi line, while it was significantly decreased in the #10.3 line (Fig. 4). All three transcripts appeared similarly reduced in the #10.3 line, whereas the ABI5 transcript level was not affected (Fig. 4). It is thus possible to trigger a specific reduction of the expression of three bZIP homologous genes by an RNAi approach using a single transgene. Fig. 4. View largeDownload slide Endogenous EEL, AREB3, AtbZIP67, and ABI5 expression in siliques from homozygous RNAi lines. One μg total RNA of 10 DAP siliques from the #10.3, #17.3, and #13 RNAi T3 homozygous lines was used to assay the endogenous expression of the EEL, AREB3, and AtbZIP67 genes by semi-quantitative RT-PCR. The primers used sit from both sides of an intron, the expected size for cDNA amplification is shown by an arrow and cloned cDNA was amplified as a control. ABI5 expression level was assayed in 14 DAP siliques from the #10.3 line and did not show any significant decrease compared to the wild-type (Col0). The elongation factor EF1α is used as an internal control for sample calibration. Fig. 4. View largeDownload slide Endogenous EEL, AREB3, AtbZIP67, and ABI5 expression in siliques from homozygous RNAi lines. One μg total RNA of 10 DAP siliques from the #10.3, #17.3, and #13 RNAi T3 homozygous lines was used to assay the endogenous expression of the EEL, AREB3, and AtbZIP67 genes by semi-quantitative RT-PCR. The primers used sit from both sides of an intron, the expected size for cDNA amplification is shown by an arrow and cloned cDNA was amplified as a control. ABI5 expression level was assayed in 14 DAP siliques from the #10.3 line and did not show any significant decrease compared to the wild-type (Col0). The elongation factor EF1α is used as an internal control for sample calibration. No macroscopic defects could be observed in the seeds of the #10.3 line. The ABA sensitivity of the T3 RNAi homozygous lines #10.3 and #17.3 is similar to the Col0 wild-type. These RNAi lines do not seem to be either hypersensitive or insensitive to ABA. Moreover, the above-mentioned germination tests were performed on seeds after a few weeks of drying (typically between 3–6 weeks) and the germination rate was always similar to the wild type, indicating that desiccation tolerance was not drastically affected as in the abi3 severe alleles. To get an insight of the molecular targets of these RNAi lines, the expression of the AtEm1 gene was assayed using semi-quantitative RT-PCR on total RNA extracted from 14 DAP #10.3 and #17.3 T3 RNAi seeds and no significant difference could be detected compared with the wild-type seeds (not shown). In addition to this, RNAi T3 homozygous lines #10.3 and #17.3 were crossed to the AtEm1∷GUS reporter line and no ectopic or precocious GUS staining (assayed at 10 and 12 DAP, not shown) could be detected in the F1 progeny of the cross, showing that there is no dominant effect on the regulation of AtEm1. By contrast with the the eel mutant, there is no earlier or stronger AtEm1 expression in the RNAi seeds. This difference can be either due to residual expression of the EEL transcript or antagonistic effect of the AtbZIP67 and AREB3 expression loss. However, a more extended molecular characterization would need to be performed to assay other possible molecular targets. Discussion GFP-bZIP pattern in planta This work demonstrates the feasibility of the detection of GFP-tagged bZIP proteins from the ABI5 family in embryos during seed maturation. GFP fusions can thus be detected even during the latest stages of seed maturation, despite the fragility of the tissues and the background fluorescence caused by the storage products. As shown with the ABI5 fusion, GFP-tagged proteins can be detected in the nuclei of cotyledons and axis tissues from maturing embryos. The temporal and subcellular localization of the GFP-ABI5 protein is in accordance with data using an ABI5–GUS fusion (Lopez-Molina et al., 2002). Brocard and co-workers have described an earlier ABI5 promoter activity in heart stage embryos (Brocard et al., 2002), but there are no data concerning the protein presence at this stage. The temporal expression pattern observed for the GFP-bZIP proteins is compatible with whole siliques mRNA levels (Bensmihen et al., 2002). The expression pattern of GFP-EEL is also compatible with its function in AtEm genes repression during the desiccation phase (Bensmihen et al., 2002). However, the proteins' localization should be confirmed by in situ immunolocalization with specific antibodies. It was observed that GFP-EEL, GFP-AREB3, and GFP-AtbZIP67 are constitutively localized in the nuclei of the embryo during maturation. Lopez-Molina and co-workers have already described a nuclear localization for an ABI5-GUS protein in dry seeds (Lopez-Molina et al., 2002). This study shows that other bZIP of the ABI5 family are constitutively located in the nucleus during embryo maturation, with no evidence for developmentally-related relocalization. Nuclear location is frequent for bZIP transcription factors, as the basic domain has been shown to act as a nuclear localization signal (van der Krol and Chua, 1991). However, some bZIP are retained in the cytoplasm (Igarashi et al., 2001) and can be targeted to the nucleus in response to external signals such as light (Kircher et al., 1999). EEL, AREB3, and AtbZIP67 display a large overlapping localization pattern in the embryo during the maturation phase. As these bZIP were shown to form heterodimers in vitro (Kim et al., 2002), they are likely to interact in vivo and perform some of their functions as heterodimers. The possibility of combinatorial action increases the range of possible functions for this set of bZIP. Because of their sequence similarity and overlapping expression patterns EEL, AREB3, and AtbZIP67 proteins may also have partially overlapping functions. RNAi approach Using a single RNAi transgene was successful in obtaining a significant reduction in the expression of the three genes. The efficiency of the silencing seems similar on the three targeted genes, since lines such as #10.3 seem to affect the EEL, AREB3, and AtbZIP67 endogenous transcripts in the same manner. This effect is also highly specific since the ABI5 transcript is not affected in these RNAi lines, whereas ABI5 is strongly homologous to EEL, AREB3, and especially AtbZIP67 (Bensmihen et al., 2002). This study demonstrates that it is possible to reduce the expression of the three homologous genes in a specific manner using a single transgene. Approaches using transgenes expressing sense chimeric constructs have already been described in tomato and tobacco to reduce the expression of two or three unrelated genes at the same time (Seymour et al., 1993; Abbott et al., 2002). Silencing of two independent genes, using the pHellsgate 8 vector to trigger intron-spliced hairpin RNA (ihpRNA) expression in Arabidopsis, has also been reported (Helliwell et al., 2002). But in each case, an average of 300 bp to 1000 bp of every targeted sequence has been used for the silencing construct and no homologous genes were targeted. This is the first time that the use of a single transgene with such short fragments has been shown to trigger multiple and specific gene silencing for homologous genes. As mentioned by others (Abbott et al., 2002; Helliwell et al., 2002), this strategy is faster than combining independent single gene silencing constructs or T-DNA insertion mutants. Even if the loss of the three bZIP expression is not complete, it might be sufficient to trigger alteration in the regulation of specific target genes expressed during seed maturation. Indeed, studies using RNAi approaches have shown that a range of severity in phenotypes could arise in different lines (Chuang and Meyerowitz, 2000), depending on the efficiency of the silencing. It is therefore possible that an intermediate silencing could still provide an intermediate phenotype in the RNAi progeny. It cannot be excluded that defects in seed maturation (such as loss of seed viability) have prevented the isolation of lines with a stronger silencing effect. As the efficiency of the silencing is related to the number of double-stranded RNA expressed (Fire et al., 1998), the approach could be improved by using a strong seed-specific promoter to drive the expression of the RNAi construct. Alternatively, greater silencing efficiency could be achieved by increasing the length of the fragments used for the RNAi constructs, as longer fragments would provide more probes to initiate the RNAi degradation process (Waterhouse and Helliwell, 2003). Since the level of each of these three bZIP transcripts is significantly reduced in some of these RNAi lines, although the expression level of AtEm1 did not seem to be affected, it will be interesting to look for other misregulated target genes of the LEA and seed storage proteins classes. This approach would give important information on the putative functions of bZIP transcription factors of the ABI5 family in the regulation of seed maturation in Arabidopsis thaliana. † Present address: Laboratoire DRDC/PCV, UMR CEA-CNRS 5168-INRA 1200-UJF, CEA, 17 rue des Martyrs, bât. C2, 38054 Grenoble Cedex 9, France. Abbreviations: ABA, abscisic acid; bp, base pair; ABI5, ABA-Insensitive 5; bZIP, basic leucine zipper; DAP, days after pollination; EEL, Enhanced Em Level; LEA, late embryogenesis abundant genes; RNAi, RNA interference. We thank C Talbot, O Catrice, S Brown, and the microscopy platform of the IFR87 in Gif-sur-Yvette (supported by the Conseil de l'Essonne) for help and access to the confocal microscope, P Waterhouse for providing the pHannibal and pART27 vectors and C Giglione for providing sequenced construct of the smRS-GFP. The pDONR201-bZIP cDNAs were cloned as part of the REGIA project by G Lambert, F Parcy, and S Bensmihen. This work was supported by the Centre National de la Recherche Scientifique and partially by the European REGIA project (grant No. QLG2-CT1999-00876). Sandra Bensmihen was supported by a grant from the French Education and Research Ministry. References Abbott JC, Barakate A, Pincon G, Legrand M, Lapierre C, Mila I, Schuch W, Halpin C. 2002. Simultaneous suppression of multiple genes by single transgenes. Down-regulation of three unrelated lignin biosynthetic genes in tobacco. Plant Physiology  128, 844–853. Google Scholar Bensmihen S, Rippa S, Lambert G, Jublot D, Pautot V, Granier F, Giraudat J, Parcy F. 2002. The homologous ABI5 and EEL transcription factors function antagonistically to fine-tune gene expression during late embryogenesis. The Plant Cell  14, 1391–1403. Google Scholar Brocard IM, Lynch TJ, Finkelstein RR. 2002. Regulation and role of the Arabidopsis abscisic acid-insensitive 5 gene in abscisic acid, sugar, and stress response. Plant Physiology  129, 1533–1543. Google Scholar Carles C, Bies-Etheve N, Aspart L, Leon-Kloosterziel KM, Koornneef M, Echeverria M, Delseny M. 2002. Regulation of Arabidopsis thaliana Em genes: role of ABI5. The Plant Journal  30, 373–383. Google Scholar Choi H, Hong J, Ha J, Kang J, Kim SY. 2000. ABFs, a family of ABA-responsive element binding factors. Journal of Biological Chemistry  275, 1723–1730. Google Scholar Chuang CF, Meyerowitz EM. 2000. Specific and heritable genetic interference by double-stranded RNA in Arabidopsis thaliana. Proceedings of the National Academy of Sciences, USA  97, 4985–4990. Google Scholar Clough SJ, Bent AF. 1998. Floral dip: a simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. The Plant Journal  16, 735–743. Google Scholar Davis SJ, Vierstra RD. 1998. Soluble, highly fluorescent variants of green fluorescent protein (GFP) for use in higher plants. Plant Molecular Biology  36, 521–528. Google Scholar Finkelstein RR, Lynch TJ. 2000. The Arabidopsis abscisic acid response gene ABI5 encodes a basic leucine zipper transcription factor. The Plant Cell  12, 599–610. Google Scholar Fire A, Xu S, Montgomery MK, Kostas SA, Driver SE, Mello CC. 1998. Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature  391, 806–811. Google Scholar Goldberg RB, de Paiva G, Yadegari R. 1994. Plant embryogenesis: zygote to seed. Science  266, 605–614. Google Scholar Helliwell C, Wesley V, Wielopolska A, Waterhouse P. 2002. High-throughput vectors for efficient gene silencing in plants. Plant Functional Genomics  29, 1217–1225. Google Scholar Hurst HC. 1995. Transcription factors 1: bZIP proteins. Protein Profile  2, 101–168. Google Scholar Igarashi D, Ishida S, Fukazawa J, Takahashi Y. 2001. 14-3-3 proteins regulate intracellular localization of the bZIP transcriptional activator RSG. The Plant Cell  13, 2483–2497. Google Scholar Jakoby M, Weisshaar B, Droege-Laser W, Vicente-Carbajosa J, Tiedemann J, Kroj T, Parcy F. 2002. BZIP transcription factors in Arabidopsis. Trends in Plant Science  7, 106–111. Google Scholar Kim SY, Ma J, Perret P, Li Z, Thomas TL. 2002. Arabidopsis ABI5 subfamily members have distinct DNA-binding and transcriptional activities. Plant Physiology  130, 688–697. Google Scholar Kircher S, Wellmer F, Nick P, Rugner A, Schafer E, Harter K. 1999. Nuclear import of the parsley bZIP transcription factor CPRF2 is regulated by phytochrome photoreceptors. Journal of Cell Biology  144, 201–211. Google Scholar Liscum E, Reed JW. 2002. Genetics of Aux/IAA and ARF action in plant growth and development. Plant Molecular Biology  49, 387–400. Google Scholar Lopez-Molina L, Chua NH. 2000. A null mutation in a bZIP factor confers ABA-insensitivity in Arabidopsis thaliana. Plant Cell Physiology  41, 541–547. Google Scholar Lopez-Molina L, Mongrand S, McLachlin DT, Chait BT, Chua NH. 2002. ABI5 acts downstream of ABI3 to execute an ABA-dependent growth arrest during germination. The Plant Journal  32, 317–328. Google Scholar Messenguy F, Dubois E. 2003. Role of MADS box proteins and their cofactors in combinatorial control of gene expression and cell development. Gene  316, 1–21. Google Scholar Parcy F, Valon C, Raynal M, Gaubier-Comella P, Delseny M, Giraudat J. 1994. Regulation of gene expression programs during Arabidopsis seed development: roles of the ABI3 locus and of endogenous abscisic acid. The Plant Cell  6, 1567–1582. Google Scholar Pelaz S, Ditta GS, Baumann E, Wisman E, Yanofsky MF. 2000. B and C floral organ identity functions require SEPALLATA MADS-box genes. Nature  405, 200–203. Google Scholar Riechmann JL, Heard J, Martin G, et al. 2000. Arabidopsis transcription factors: genome-wide comparative analysis among eukaryotes. Science  290, 2105–2110. Google Scholar Seymour GB, Fray RG, Hill P, Tucker GA. 1993. Down-regulation of two non-homologous endogenous tomato genes with a single chimaeric sense gene construct. Plant Molecular Biology  23, 1–9. Google Scholar Stracke R, Werber M, Weisshaar B. 2001. The R2R3-MYB gene family in Arabidopsis thaliana. Current Opinion in Plant Biology  4, 447–456. Google Scholar van der Krol AR, Chua NH. 1991. The basic domain of plant B-ZIP proteins facilitates import of a reporter protein into plant nuclei. The Plant Cell  3, 667–675. Google Scholar Voinnet O, Vain P, Angell S, Baulcombe DC. 1998. Systemic spread of sequence-specific transgene RNA degradation in plants is initiated by localized introduction of ectopic promoterless DNA. Cell  95, 177–187. Google Scholar Waterhouse PM, Helliwell CA. 2003. Exploring plant genomes by RNA-induced gene silencing. Nature Reviews Genetics  4, 29–38. Google Scholar Wesley SV, Helliwell CA, Smith NA, et al. 2001. Construct design for efficient, effective and high-throughput gene silencing in plants. The Plant Journal  27, 581–590. Google Scholar Journal of Experimental Botany, Vol. 56, No. 412, © Society for Experimental Biology 2005; all rights reserved
TI - Characterization of three homologous basic leucine zipper transcription factors (bZIP) of the ABI5 family during Arabidopsis thaliana embryo maturation
JF - Journal of Experimental Botany
DO - 10.1093/jxb/eri050
DA - 2005-01-10
UR - https://www.deepdyve.com/lp/oxford-university-press/characterization-of-three-homologous-basic-leucine-zipper-1KkYCzFeps
SP - 597
EP - 603
VL - 56
IS - 412
DP - DeepDyve
ER -