ABA is a Major Mediator of Plant Stress Response Signaling

For many years, investigators have observed that the plant hormone ABA accumulates in plant tissues, especially leaves, in response to osmotic-based stresses, including NaCl stress (Himmelbach et al., 2003; Zhu, 2002). Although other plant hormones including ethylene, salicylic acid, and jasmonic acid may also participate in various stress responses and even have interactive roles, ABA has remained the most important plant hormone controlling response and adaptation to abiotic stress.

A major mechanism by which ABA controls the plant adaptive response to osmotic/NaCl stress is thought to involve the alteration of gene expression, and many osmotic stress-responsive genes whose expression is mediated by ABA have been identified (Hoth et al., 2002; Seki et al., 2002). Confirmation of the control of gene expression in adaptation to stress has been provided by studies that show the effects of genes controlling ABA sensitivity on gene expression (Xiong and Zhu, 2003). The ABA-insensitive mutants abi1 and abi2 dramatically affect the ability of ABA to induce gene expression changes. Although important gene expression changes induced by stress are induced by ABA, it is now well accepted that stress-induced gene expression changes in plants are rather complex; they can be both dependent and independent of ABA mediation (Grillo et al., 1995; Ingram and Bartels, 1996; Kurkela and Franck, 1990; Nordin et al., 1991).

In addition to genes that affect ABA sensitivity of target gene expression, genes that encode proteins controlling ABA biosynthesis have been identified and found to be themselves controlled by stress exposure (Xiong and Zhu, 2003). Therefore, the long-standing observation of a stress-induced increase in tissue ABA levels appears to be at least partially the result of transcriptional regulation. The main biosynthetic pathway for ABA is initiated in plants by the conversion of zeaxanthin to violaxanthin by the enzyme zeaxanthin epoxidase (ZEP) (Audran et al., 1998). An important and perhaps rate limiting step is the oxidative cleavage of neoxanthin by 9-cis epoxycarotenoid deoxygenase (NCED). These enzymes and the ABA aldehyde oxidase (AAO) and the molybdenum cofactor sulfurylase (MCSU) have all been shown to be transcriptionally upregulated by drought or salt stress (Audran et al., 1998; Iuchi et al., 2000; Seo et al., 2000; Xiong et al., 2001a, 2002a). Since, genes encoding enzymes for several steps in the ABA biosynthetic pathway may be also transcriptionally activated by ABA itself, through a possible Ca ⁺⁺ /phosphorylation signal cascade, ABA may participate in a feed-forward loop to amplify ABA-mediated stress responses (Xiong et al., 2001a,b, 2002a,b; Zhao et al., 2001).

Besides controlling phenotype by way of modulating the activity of TFs such as AB1–3, AB1–4, and AB1–5, it is clear that ABA is involved in controlling many aspects of accessing the mediation between the environment and appropriate response information encoded within plant genomes (Hoth et al., 2002). The SAD1 gene describes a mutant with phenotypic alterations involving ABA that are manifested through processes at the posttranscriptional or RNA metabolism level of control and several other genes in this function have been described (ABH1, CPL3, CPL1, HYL-1) (Hoth et al., 2002; Hugouvieux et al., 2002; Koiwa et al., 2004; Xiong et al., 2001a, 2002a). The SAD1 gene encodes an SM-like SnRNA protein that when mutated conveys supersensitive responses to drought and ABA (Xiong et al., 2001a, 2002a). The HYL1 gene also mediates ABA responses by causing hyperaccumulation of AB15 mRNA and protein [HYL1 encodes a double stranded (dsRNA) bind protein and probably affects ABA] responses by processing or stabilization of RNA (Lu et al., 2002). The Fiery-2 and AtCPL1 are alleles of the gene that encodes RNA polymerase II carboxy-terminal domain phosphataselike proteins that also control RNA polymerase II activity to specifically mediate the efficiency of ABA/stress responses (Koiwa et al., 2002; Xiong et al., 2002b). In addition, mRNA cap structure and transcript maturation are also involved in ABA control of stress responses as revealed by the ABA hypersensitive 1-cap-binding protein2 (Hugouvieux et al., 2001, 2002). It is likely that several more genes involving ABA-mediated phenotypes will be discovered that control many aspects of transcript processing and translation. Also, given the rapidly rising importance of unusual epigenetic control processes such as miRNA, SiRNA, DNA, and chromatin protein modifications (e.g., methylation or acetylation), it is becoming clear that there are epigenetic coding systems in plants that encode dynamic information guiding development and environmental responses beyond that directed by the encoded information of the DNA sequence and the central dogma of transcription and translation (Wada et al., 2004). These information-encoding systems will certainly play important roles in the function of plant stress responses including those mediated by ABA (Himmelbach et al., 2003).

It is clear now that there are many signaling components that control the adaptive responses of plants to salt/osmotic stress. Furthermore, the identification of a number of these components has clearly shown that stress-responsive signal pathways overlap between many different stresses (Hasegawa et al., 2000a; Xiong et al., 2002b). This is a clear indication that we will need to identify many more components of stress signaling in order to understand their interrelationships and how they cooperate to mediate responses to multiple stresses that often occur together in the environment. Several important technical advances are allowing the opportunity to accomplish this formidable task.

The completion of the sequencing of the Arabidopsis genome has paved the way for both forward and reverse genetic screens aimed at the identification of mutants with a myriad of phenotypes including many with altered stress tolerances. In addition, the collection of tagged insertion mutants available from the SALK Center that represent mutations in nearly all expressed genes of Arabidopsis has so simplified the task of identifying which gene is responsible for any particular mutant phenotype, that functional genomics analysis has become a tool now available to investigators with less experience in molecular genetics. In fact, the tool of insertion tagging has been surprisingly successful and is now rapidly allowing further dissection of signal pathways through the production of many more mutant phenotypes including second site mutations that suppress or enhance original phenotypes (Rus et al., 2004). The emerging new phenotypes and the genes responsible for them will greatly increase our understanding of the genetic basis of adaptation of plants to stress environments.