|
After the biochemical pathways producing storage lipids in oilseed species like Brassica napus or Arabidopsis thaliana were largely described (Voelker and Kinney, 2001), the elucidation of the regulation of oil synthesis has become a major challenge (Ohlrogge and Jaworski, 1997). Complementary studies carried out on developing seeds and/or embryos have established that the biosynthetic pathways for fatty acids and lipids are largely regulated at the level of transcription (Fawcett et al., 1994; O’Hara et al., 2002; Baud and Graham, 2006). In A. thaliana, microarray analyses have been used to characterize the ‘contrapuntal’ or differential timing of the expression of genes involved in fatty-acid biosynthesis and lipid metabolism during seed maturation (Ruuska et al., 2002). Genes related to the biosynthesis and storage of triacylglycerols (TAGs) show several distinct temporal expression patterns. For instance, a number of genes encoding fatty acid synthesis enzymes display a bell-shaped pattern of expression at the onset of the maturation phase, whereas the expression of fatty-acid modifying enzymes and oleosins increases later (Baud and Lepiniec, 2009). The observation that the mRNAs of several genes encoding enzymes of the fatty acid biosynthetic pathway accumulate in a coordinated manner strongly suggests that those members of the pathway are co-regulated, and presumably share common cis- and trans-regulatory elements. This hypothesis has been strengthened by the isolation and characterization of the WRINKLED1 (WRI1) transcription factor (Focks and Benning, 1998; Cernac and Benning, 2004; Masaki et al., 2005). WRI1 is a direct target of LEAFY COTYLEDON2 (LEC2) that specifies the regulatory action of this master regulator of seed maturation towards fatty acid metabolism (Baud et al., 2007a). Putative targets of WRI1 encode enzymes of late glycolysis, the fatty acid synthesis pathway, and the biotin and lipoic acid biosynthetic pathways (Ruuska et al., 2002; Baud et al., 2007a).
WRI1 is a member of the APETALA2/ethylene-responsive element binding (AP2/EREBP) proteins, one of the largest transcription factor families in A. thaliana (Riechmann et al., 2000). This family is best characterized by a common AP2 domain of about 60 amino acids that is important for DNA binding (Jofuku et al., 1994; Okamuro et al., 1997; Riechmann and Meyerowitz, 1998; Sakuma et al., 2002). While the AP2/EREBP transcription factor family is unique to plants, proteins containing homologues of the AP2 domain have been identified in cyanobacteria, ciliates and viruses, where they are predicted to function as mobile endonucleases (Magnani et al., 2004; Wessler, 2005; Shigyo et al., 2006). Based on the presence of the conserved AP2-like domains, 147 proteins were identified as belonging to the AP2/EREBP family in A. thaliana (Feng et al., 2005). The AP2/EREBP family members are classified in groups based on the number of AP2 domains and the presence of other domains. The EREBP family (122 members), which has a single AP2 domain, is composed of two subgroups, namely the ERF (65 members) and DREB subfamilies (57 members). Proteins of the RAV family (six members) contain one AP2 domain and a second DNA binding domain (the B3-like domain). The AP2 family (19 members), which has two AP2 domains, is further divided into two monophyletic groups, the APETALA2-like and AINTEGUMENTA (ANT)-like groups. Depending on the publications, the WRI1-like group, composed of WRI1, At2g41710, At1g16060 and At1g79700, is either presented as a third monophyletic group of the AP2 family (Nole-Wilson et al., 2005) or as a part of the ANT-like group (Shigyo et al., 2006).
DNA binding specificity has been studied for proteins that contain a single AP2 repeat. Proteins of the ERF subfamily bind to ethylene response elements (ERE) or GCC boxes. These boxes consist of 11-bp sequences (TAAGAGGCCGCC) exhibiting a GCCGCC core required for protein binding (Ohme-Takagi and Shinshi, 1995; Fujimoto et al., 2000; Song et al., 2005). The solution structure of the AP2 domain of AtERF1 bound to the GCC box has been determined (Allen et al., 1998). Proteins of the DREB subfamily bind to dehydration response elements (DREs)/C-repeats containing the core sequence CCGAC (Baker et al., 1994). The single AP2 repeat of the RAV1 transcription factor has been shown to bind to a CAACA motif, whereas the B3 domain binds to a CACCTG sequence (Kagaya et al., 1999). Little is known about the interactions between members of the AP2 family of proteins and DNA. A consensus binding sequence of the AINTEGUMENTA (ANT) protein determined in vitro is 5′-gCAC(A/G)N(A/T)TcCC(a/g)ANG(c/t)-3′ (Nole-Wilson and Krizek, 2000). This consensus is much longer than sites recognized by proteins containing a single AP2 repeat. Each AP2 repeat of ANT contacts juxtaposed subsites within the consensus sequence, demonstrating that ANT utilizes a mode of DNA recognition distinct from that used by proteins containing a single AP2 domain (Krizek, 2003). To date, given the lack of information about the binding properties of proteins of the AP2-like and WRI1-like groups, it has not been possible to determine whether all of the transcription factors of the AP2 family share a similar mode of DNA recognition.
In this paper, it is shown that WRI1 can be used in a seed-specific manner to enhance the transcription level of glycolytic and fatty acid biosynthetic genes in tissues where these genes are already expressed. Consistent with this, it is demonstrated that WRI1 is able to regulate in planta the activity of the BCCP2 and PKp-β1 promoters. In addition, electrophoretic mobility-shift assays and yeast one-hybrid experiments show that WRI1 is able to interact with the BCCP2 promoter. These results strongly suggest that WRI1, which is a limiting factor of lipogenic gene expression in seeds, directly induces the transcriptional activation of these genes at the onset of the maturation phase. Functional dissections of the PKp-β1 and BCCP2 promoters allow isolating putative binding sites to be present in several target promoters of WRI1. Finally, the modification of the consensus sequence thus isolated modulates the strength of the interaction between WRI1 and the BCCP2 promoter, both in vitro and in yeast, as well as the activity of the promoter in planta. Taken together, these results provide original insights into our understanding of the transcriptional activation of the glycolytic and fatty acid biosynthetic pathways in A. thaliana, and they pave the way for further analyses aimed at elucidating the regulatory network controlling oil accumulation in seeds.
|
