Abscisic acid (ABA) signalling is involved in several abiotic stress responses, germination and seedling development. ABF2 is a transcription factor that binds to ABA-responsive promoter elements. ABF2 interacts with ARIA (Arm-repeat protein interacting with ABF2), which has a BTB/POZ domain (Kim et al. 2004a,b; Fig. 1).
Aria mutant seedlings show reduced sensitivity to ABA during germination and root growth. This suggests that ARIA functions as a positive regulator of ABA signalling, and accordingly ARIA mRNA is upregulated by ABA
Fig. 2 Functions and mechanisms of plant Arm-repeat protein signalling. A U-box/Arm ► proteins function as E3 ubiquitin ligases during a variety of physiological and developmental processes; interacting transmembrane receptors are known for NtPUB4 and ARC1 only. B UPL3/KAKTUS may respond to gibberellin to degrade inhibitors of trichome branching, DNA replication and cell expansion. C ARIA binds to ABF transcription factors that regulate stress- and glucose-responses in early development; they may also be part of CULLIN3-containing ubiquitin ligase complexes. D ARABIDILLO-1 and -2 are postulated to act as part of an SCF ubiquitin ligase complex to promote lateral root development. E MRH2 contains a kinesin domain and regulates microtubule-mediated root hair morphogenesis. TIO is required for cytokinesis and may interact with microtubules, possibly via phosphorylation
(Kim et al. 2004a). However, abf2 mutants are not ABA-insensitive (Kim et al. 2004b). Both abf2 and aria mutants show reduced sensitivity to glucose (Kim et al. 2004a). Moreover, mutants in ABF2-related transcription factors, ABF3 and ABF4, are ABA-insensitive, suggesting that ARIA may regulate all three ABFs (Kim et al. 2004b; Fig. 2).
ARIA-overexpressing plants are hypersensitive to ABA, as would be predicted from the knockout phenotype. However, they also show hypersensitiv-ity to osmotic stress during germination and salt-tolerance in later life (Kim et al. 2004a); both of these phenotypes could arise from ARIA interacting with ABF3 and ABF4 (Kim et al. 2004b).
Unlike ABF2, which is exclusively nuclear, an ARIA-GFP fusion is found both in the nucleus and at the cell periphery (Kim et al. 2004a), suggesting that ARIA could have non-nuclear functions. There is no evidence as yet that ARIA protein activity or localisation is regulated directly by ABA. In addition, ARIA affects only a subset of ABA-dependent processes (Kim et al. 2004a). Arabidopsis possesses one ARIA-related gene, and thus the relatively mild phenotypes seen in the aria mutant may be due to redundancy.
Although the BTB/POZ domain was originally characterised in transcription factors, it is also present in ubiquitin ligases (Gingerich et al. 2005). Thus, ARIA could degrade target proteins in addition to directly regulating ABF-mediated transcription (Fig. 2). In general, aria and abf mutants have similar phenotypes, suggesting ARIA does not degrade ABFs themselves. Interestingly, other Arm proteins may be involved in degradation responses to stress: in mangrove, bg55 mRNA (encoding a U-box/Arm protein) is upregulated by salt stress (Banzai and Karube 2002).
Was this article helpful?