The Role of Stress-Responsive Transcription Factors in Modulating Abiotic Stress Tolerance in Plants

: Abiotic stresses, such as drought, high temperature, and salinity, a ﬀ ect plant growth and productivity. Furthermore, global climate change may increase the frequency and severity of abiotic stresses, suggesting that development of varieties with improved stress tolerance is critical for future sustainable crop production. Improving stress tolerance requires a detailed understanding of the hormone signaling and transcriptional pathways involved in stress responses. Abscisic acid (ABA) and jasmonic acid (JA) are key stress-response hormones in plants, and some stress-responsive transcription factors such as ABFs and MYCs function as direct components of ABA and JA signaling, playing a pivotal role in plant tolerance to abiotic stress. In addition, extensive studies have identiﬁed other stress-responsive transcription factors belonging to the NAC, AP2 / ERF, MYB, and WRKY families that mediate plant response and tolerance to abiotic stress. These suggest that transcriptional regulation of stress-responsive genes is an essential step to determine the mechanisms underlying plant stress responses and tolerance to abiotic stress, and that these transcription factors may be important targets for development of crops with enhanced abiotic stress tolerance. In this review, we brieﬂy describe the mechanisms underlying plant abiotic stress responses, focusing on ABA and JA metabolism and signaling pathways. We then summarize the diverse array of transcription factors involved in plant responses to abiotic stress, while noting their potential applications for improvement of stress tolerance. Functional validation under ﬁeld conditions will expand potential applications of stress-responsive transcription factors for improvement of abiotic stress tolerance in


Introduction
Stress-responsive transcription factors play essential roles in abiotic stress responses and stress tolerance [1,2]. Therefore, these stress-responsive transcription factors may be important targets for developing crops with enhanced abiotic stress tolerance [3]. Plant stress hormones, such as abscisic acid (ABA) and jasmonic acid (JA), regulate plant abiotic stress responses [4,5]. ABA signaling pathways activate target transcription factors, such as the basic leucine zipper (bZIP) ABA-responsive element binding factor (ABF) transcription factors, and JA signaling pathways activate the basic helix-loop-helix (bHLH) MYC transcription factors. These transcription factors control ABA-and JA-dependent expression of stress-responsive genes, as demonstrated using knock-out and overexpression systems [6][7][8]. In addition, experimental and computational approaches have identified other stress-responsive transcription factors belonging to the NAC, AP2/ERF, MYB, and WRKY families, which are not direct components of ABA and JA signaling pathways but are essential for plant abiotic stress responses. In this review, we describe the roles of ABA and JA metabolism and signaling Figure 1. A schematic of ABA signaling pathway. In response to ABA, ABA receptor PYR1/PYL/RCAR forms a complex with PP2Cs, and activates SnRKs, the kinases that determine the activity of ABF transcription factors. SnRK-mediated phosphorylation of ABF promotes expression of ABA-responsive genes and ABA response.
The pivotal functions of ABFs in the ABA response and abiotic stress tolerance have been well documented in Arabidopsis thaliana. The transcription of ABF1, 2, 3, and 4 is strongly upregulated by ABA or abiotic stress, and overexpression of ABF3 and 4 promotes the expression of ABA-responsive genes, such ABA-INSENSITIVE 1 (ABI1) and ABI2, which encode PP2Cs, and whose mutations cause defects in ABA-dependent stomatal closure [35][36][37]. Furthermore, overexpression of ABF3 and 4 significantly improves drought tolerance. For example, most wild-type plants withered by approximately 10 days of drought stress, whereas ABF3-and 4-overexpressing transgenic plants survived the drought stress [36]. In contrast to the ABF3-and 4-overexpressing plants, the abf2 abf3 abf4 triple mutant exhibited reduced drought tolerance compared to wild-type plants, and large-scale transcriptome analysis showed that expression of stress-responsive genes is impaired in the triple mutant [27]. These studies indicate that ABFs play an essential role in the plant response to abiotic stress. Moreover, ABF3 and 4 are predominantly expressed in roots, suggesting that ABF3 and 4 are involved in ABA response in roots [27,36]. Despite the functional redundancy of ABFs in ABA response and stress tolerance, it is likely that their functions do not completely overlap [36,38].
ABFs mediate stress-induced changes in plant development. For example, abiotic stress accelerates flowering and shortens the plant life cycle. A recent study by Hwang et al. revealed that ABF3 and 4 regulate flowering time by controlling the expression of the floral integrator SUPPRESSOR OF OVEREXPRESSION OF CONSTANS 1 (SOC1) [38]. Mutant plants that lack expression of ABF3 and 4 exhibit reduced expression of SOC1 and showed an ABA-insensitive late flowering phenotype, whereas transgenic plants ectopically expressing ABF3 or 4 displayed an early flowering phenotype. Together with the result that SOC1 is a direct downstream target of ABF3 and 4 transcription factors, these findings suggest that stress-responsive ABFs mediate ABA-dependent modulation of plant growth and development under stress conditions.
The function of ABFs in abiotic stress tolerance is conserved in plants. For example, ABA and abiotic stress upregulate the Fagopyrum tataricum FtbZIP5 and Poncirus trifoliata PtrABF bZIP transcription factors. FtbZIP5-overexpressing Arabidopsis plants exhibited improved tolerance to drought, salinity, and oxidative stress by modulating the antioxidant system [39]. Similar to FtbZIP5, Figure 1. A schematic of ABA signaling pathway. In response to ABA, ABA receptor PYR1/PYL/RCAR forms a complex with PP2Cs, and activates SnRKs, the kinases that determine the activity of ABF transcription factors. SnRK-mediated phosphorylation of ABF promotes expression of ABA-responsive genes and ABA response.
The pivotal functions of ABFs in the ABA response and abiotic stress tolerance have been well documented in Arabidopsis thaliana. The transcription of ABF1, 2, 3, and 4 is strongly upregulated by ABA or abiotic stress, and overexpression of ABF3 and 4 promotes the expression of ABA-responsive genes, such ABA-INSENSITIVE 1 (ABI1) and ABI2, which encode PP2Cs, and whose mutations cause defects in ABA-dependent stomatal closure [35][36][37]. Furthermore, overexpression of ABF3 and 4 significantly improves drought tolerance. For example, most wild-type plants withered by approximately 10 days of drought stress, whereas ABF3and 4-overexpressing transgenic plants survived the drought stress [36]. In contrast to the ABF3and 4-overexpressing plants, the abf2 abf3 abf4 triple mutant exhibited reduced drought tolerance compared to wild-type plants, and large-scale transcriptome analysis showed that expression of stress-responsive genes is impaired in the triple mutant [27]. These studies indicate that ABFs play an essential role in the plant response to abiotic stress. Moreover, ABF3 and 4 are predominantly expressed in roots, suggesting that ABF3 and 4 are involved in ABA response in roots [27,36] . This indicates functional conservation of stress-responsive bZIPs in plants, suggesting that modulation of the stress-responsive bZIPs could be a good strategy for development of crops with improved stress tolerance. For example, ABA and abiotic stress induce the expression of rice (Oryza sativa) OsbZIP23 [41] and transactivation assays showed that OsbZIP23 is a transcriptional activator. Overexpression of OsbZIP23 increased sensitivity to ABA and improved the tolerance to abiotic stress, including drought and salinity.
Other stress-responsive rice bZIPs showed similar functions in abiotic stress tolerance to that of OsbZIP23. For example, OsbZIP12 and OsbZIP71 conferred drought tolerance by increasing sensitivity to ABA, and OsbZIP46 and OsbZIP72 improved drought tolerance in rice and Arabidopsis [36,42-46]. ZmbZIP4 is a stress-responsive maize (Zea mays) bZIP transcription factor with high sequence similarity to OsbZIP23 [47]. Similar to OsbZIP23, overexpression of ZmbZIP4 increased ABA levels and enhanced abiotic stress tolerance by activating the transcription of stress-responsive genes. Interestingly, ZmbZIP4 also affects expression of genes involved in root development in maize, such as ZmLRP1, ZmSCR, ZmIAAs, and ZmARFs, and ZmbZIP4-overexpressing transgenic plants formed longer roots with an increased number of lateral roots. These results suggest that ZmbZIP4 positively regulates plant abiotic stress responses and is involved in root development in maize [47].

JA Metabolism
JA regulates plant responses to environmental stresses and modulates development under stress conditions [48,49]. JA, a cyclopentane fatty acid, and was initially isolated as a methyl ester from Jasminum grandiflorum. JA is biosynthesized from linolenic acid via the octadecanoid pathway involving a variety of enzymatic components, such as lipoxygenase (LOX), allene oxide synthase (AOS), allene oxide cyclase (AOC), and 12-oxo-PDA reductase (OPR) [50,51]. The free acid form of JA, which is produced through the octadecanoid pathway, is further metabolized into the JA-isoleucine conjugate (JA-Ile) or methyl jasmonate (MeJA) through the activity of jasmonate-amido synthetase 1 (JAR1) and jasmonate methyl transferase (JMT), respectively [52,53]. JA-Ile is an active form of JA that interacts with JA receptors to promote the JA-dependent stress response [54]. In response to abiotic stress, the expression of JA biosynthesis genes is dynamically upregulated, leading to increases in endogenous JA levels. For example, cold temperature stress rapidly activated the JA response by inducing expression of JA biosynthesis genes such as LOX, AOS, and AOC in Arabidopsis and rice [55,56]. Furthermore, exogenous JA treatments improved freezing tolerance, and the JA-deficient mutants lox2 and aos exhibited reduced tolerance to freezing compared to wild-type plants [55]. In addition, transgenic Arabidopsis plants overexpressing wheat (Triticum aestivum) AOC1 (TaAOC1) showed enhanced tolerance to salt stress by activating JA biosynthesis [57,58]. These studies support the view that JA is a key hormone mediating plant responses and tolerance to abiotic stress, indicating that JA metabolism is a key step controlling the JA response in plants. Similar to ABA, JA transport remains largely unknown. A radioisotopic labeling experiment suggested that JA is transported both through phloem and xylem [59]. However, a recent study using a micrografting experiment suggested shoot-to-root JA transport mainly though phloem, and this finding is partially supported by phloem-specific accumulation of JA biosynthetic enzymes such as LOX and AOS [60,61].

JA Signaling and MYC Transcription Factors
Early studies on JA revealed that the expression of genes involved in plant defenses is rapidly and dynamically regulated by exogenous JA treatment, indicating the existence of a JA-specific signaling pathway and the integral role of JA in modulating stress-related transcription networks [62,63]. In Arabidopsis, identification of JA signaling components, including the JA receptor CORONATINE INSENSITIVE 1 (COI1) and the JA signaling repressors JASMONATE ZIM-DOMAIN proteins (JAZs), have extended our understanding of the JA signaling pathway and establishment of JA-dependent transcription. In response to abiotic stress, JA, especially JA-Ile, activates JA signaling through interaction with COI1. This interaction induces 26S proteasome-mediated proteolysis of JAZs. In the absence of JA, the activity of the MYC2 transcription factor, which governs the expression of JA-dependent stress-responsive genes, is suppressed through its interaction with JAZ proteins. The degradation of JAZs liberates MYC2, which then activates the expression JA-responsive genes [64][65][66][67] (Figure 2). This indicates a pivotal role of MYC2 in the JA response, and the finding that expression of MYC2 is regulated by diverse abiotic stresses, including drought and salinity, suggests that MYC2 is involved in the plant response and tolerance to abiotic stress [68][69][70]. MYC2 was identified in a mutant screen for insensitivity to exogenous JA, and further characterization revealed that MYC2 is a key component of the JA signaling pathway [71]. MYC2 contains a conserved bHLH domain required for the formation of homo-or hetero-dimers with other JA-responsive MYCs such as MYC3 and MYC4 [72]. The basic region of the MYC2 protein is responsible for its interaction with DNA (G-box; 5 -CACGTG-3 ) [73]. Expression of MYC2 is involved in the plant response and tolerance to abiotic stress such as oxidative stress. Ascorbate is a major reactive oxygen species (ROS) scavenger in plants [74]. Exogenous JA treatment increases ascorbate levels by inducing ascorbate biosynthesis genes such as DEHYDROASCORBATE REDUCTASE and ASCORBATE PEROXIDASE, and knock-out mutant plants that lack MYC activity display reduced expression of genes involved in oxidative stress tolerance [69,75]. This suggests that MYC2 positively regulates oxidative stress tolerance, and the result that MYC2-overexpressing transgenic plants show improved tolerance to oxidative stress further supports this.
Agronomy 2020, 10, x FOR PEER REVIEW 5 of 22 DOMAIN proteins (JAZs), have extended our understanding of the JA signaling pathway and establishment of JA-dependent transcription. In response to abiotic stress, JA, especially JA-Ile, activates JA signaling through interaction with COI1. This interaction induces 26S proteasomemediated proteolysis of JAZs. In the absence of JA, the activity of the MYC2 transcription factor, which governs the expression of JA-dependent stress-responsive genes, is suppressed through its interaction with JAZ proteins. The degradation of JAZs liberates MYC2, which then activates the expression JA-responsive genes [64][65][66][67] (Figure 2). This indicates a pivotal role of MYC2 in the JA response, and the finding that expression of MYC2 is regulated by diverse abiotic stresses, including drought and salinity, suggests that MYC2 is involved in the plant response and tolerance to abiotic stress [68][69][70]. MYC2 was identified in a mutant screen for insensitivity to exogenous JA, and further characterization revealed that MYC2 is a key component of the JA signaling pathway [71]. MYC2 contains a conserved bHLH domain required for the formation of homo-or hetero-dimers with other JA-responsive MYCs such as MYC3 and MYC4 [72]. The basic region of the MYC2 protein is responsible for its interaction with DNA (G-box; 5′-CACGTG-3′) [73]. Expression of MYC2 is involved in the plant response and tolerance to abiotic stress such as oxidative stress. Ascorbate is a major reactive oxygen species (ROS) scavenger in plants [74]. Exogenous JA treatment increases ascorbate levels by inducing ascorbate biosynthesis genes such as DEHYDROASCORBATE REDUCTASE and ASCORBATE PEROXIDASE, and knock-out mutant plants that lack MYC activity display reduced expression of genes involved in oxidative stress tolerance [69,75]. This suggests that MYC2 positively regulates oxidative stress tolerance, and the result that MYC2-overexpressing transgenic plants show improved tolerance to oxidative stress further supports this. MYC transcription factors, which are characterized by the basic helix-loop-helix (bHLH) domain, belong to a subfamily of bHLH transcription factors [76,77]. The family of bHLH transcription factors is widespread in eukaryotes, and has expanded in plants [78,79]. Based on the sequence similarity of the bHLH domain, it has been suggested that Arabidopsis contains approximately 160 bHLH transcription factors, and many bHLH transcription factors, including AtbHLH17, AtbHLH68, and AtbHLH122, are involved in plant abiotic stress responses [80][81][82][83]. The function of bHLH transcription factors in abiotic stress tolerance is conserved in plants, as supported by studies showing that heterologous expression of F. tataricum FtbHLH2 and 3, and Populus MYC transcription factors, which are characterized by the basic helix-loop-helix (bHLH) domain, belong to a subfamily of bHLH transcription factors [76,77]. The family of bHLH transcription factors is widespread in eukaryotes, and has expanded in plants [78,79]. Based on the sequence similarity of the bHLH domain, it has been suggested that Arabidopsis contains approximately 160 bHLH transcription factors, and many bHLH transcription factors, including AtbHLH17, AtbHLH68, and AtbHLH122, are involved in plant abiotic stress responses [80][81][82][83]. The function of bHLH transcription factors in abiotic stress tolerance is conserved in plants, as supported by studies showing that heterologous expression of F. tataricum FtbHLH2 and 3, and Populus euphratica PebHLH35 in Arabidopsis enhanced tolerance to abiotic stresses, such as cold, drought, and oxidative stress [82,84,85].
In rice, OsbHLH148 has a similar activity to MYC2 [86]. OsbHLH148 directly interacts with JAZ proteins, suggesting that OsbHLH148 act as a JA signaling component in rice. The expression of OsbHLH148 is rapidly upregulated by JA and diverse abiotic stresses such as drought, salinity, cold, and wounding. Furthermore, transgenic rice overexpressing OsbHLH148 exhibited increased transcript levels of DEHYDRATION-RESPONSIVE ELEMENT-BINDING PROTEINS (OsDREBs), which are involved in the drought-stress response, and displayed enhanced tolerance to drought [86]. This suggests that stress-responsive bHLH transcription factors, including MYC2 and OsbHLH148, play an essential role in plant response and tolerance to abiotic stress. Indeed, studies using stress-responsive bHLH transcription factors, such as INDUCER OF CBF EXPRESSION 1 (ICE1) and ICE2, support this [55,87].
In addition, stress-responsive bHLH transcription factors, including MYC2 and OsbHLH148, are involved in ABA-mediated stress tolerance [86,88]. Expression of MYC2 and OsbHLH148 is strongly upregulated by ABA and ABA-related stresses. In addition, MYC2 promotes the expression of ABA and the dehydration-responsive gene rd22 by directly interacting with its promoter, and myc2 mutant plants show reduced sensitivity to ABA [88]. These results suggest that JA interacts with ABA to modulate stress response and tolerance, and stress-responsive bHLH transcription factors are involved in the JA-ABA crosstalk. This idea is supported by increasing evidence showing that ABA is involved in modulation of stress tolerance by stress-responsive bHLH transcription factors [89][90][91]. In addition, the results showing that activation of JA biosynthesis by jasmonic acid methyltransferase (JMT) promotes ABA production, and that the ABA receptor PYL6/RCAR9 interacts with MYC2 also support this [92,93].

NAC Transcription Factors
NAC (NO APICAL MERISTEM (NAM), ARABIDOPSIS TRANSCRIPTION ACTIVATOR FACTOR 1/2 (ATAF1/2), and CUP-SHAPED COTYLEDON 2 (CUC2)) transcription factors are plant-specific transcription factors [94,95]. Plants contain more than 100 NAC transcription factors, indicating that the NACs constitute one of the largest families of transcriptional regulators in plants. This suggests that NAC-dependent transcription networks are involved in many biological processes in plants [96,97]. NAC transcription factors contain a conserved NAC domain in the N-terminal region and a regulatory domain in the C-terminal region, which are responsible for DNA binding and transcriptional regulation, respectively [98,99]. The expression of many NACs is regulated by abiotic stress. For example, Jiang and Deyholos (2006) reported that the expression of 33 NAC genes was significantly upregulated by salt stress in Arabidopsis [100], and Fang et al. (2008) and Le et al. (2011) reported that the expression of approximately 40 NACs was upregulated in response to abiotic stresses, such as drought and salinity, in rice and soybean [101,102]. These studies suggested that stress-responsive NACs are involved in plant abiotic stress responses. Indeed, many studies of stress-responsive NACs including ANAC019, ANAC055, ANAC072, OsNAC6, OsNAC5, OsNAC9, and OsNAC10, have demonstrated this [103][104][105][106][107].
EARLY RESPONSIVE TO DEHYDRATION STRESS 1 (ERD1), encoding a chloroplast ATP-dependent protease, is involved in the response to abiotic stresses [108]. ERD1 expression is tightly regulated by abiotic stress, and the stress-responsive expression pattern of ERD1 largely depends on a CATGTG motif in its promoter. ANAC019, ANAC055, and ANAC072, which interact with the motif, are responsible for the expression of ERD1 [103], and further characterization of ANAC019, ANAC055, and ANAC072 revealed that their expression is induced by diverse abiotic stresses, such as drought and salinity, and by ABA, and overexpression of these NACs significantly increases drought tolerance. In contrast to the overexpressing plants, mutant plants that lack the activity of ANAC019, ANAC055, and ANAC072 displayed reduced tolerance to salt stress compared to wild-type control plants [109]. These findings suggest that stress-responsive NACs play a crucial role in plant tolerance to abiotic stress, and the results of other stress-responsive NACs, such as RD26 and ANAC042, support this [7,110].
The function of stress-responsive NACs is conserved in plants. Heterologous overexpression of Arabidopsis ANAC042/AtJUB1 improved abiotic stress tolerance in tomato (Solanum lycopersicum) and Arabidopsis [110,111]. In addition, transgenic Arabidopsis plants expressing stress-responsive wheat TaNAC29 exhibited enhanced tolerance to high salinity by reducing the accumulation of hydrogen peroxide [112]. This suggests that stress-responsive NACs would be good targets for improvement of abiotic stress tolerance in crops. For example, OsNAC6 is a rice NAC transcription factor, and its expression is upregulated by a variety of abiotic stresses [104]. OsNAC6-overexpressing transgenic rice showed enhanced tolerance to drought, whereas OsNAC6 knock-out mutants displayed reduced tolerance compared to wild-type control plants. This indicated that OsNAC6 positively regulates abiotic stress tolerance [112]. In addition, genome-wide transcription profiling revealed that OsNAC6 regulates the expression of nicotianamine biosynthesis genes such as NICOTIANAMINE SYNTHASE 2 (NAS2), and overexpression of NAS2 improved tolerance to drought. These findings suggest that OsNAC6 modulates plant tolerance to drought by regulating nicotianamine biosynthesis [113].
The essential role of stress-responsive OsNACs in stress tolerance was also shown in other studies using OsNAC5, OsNAC9, and OsNAC10. Similar to OsNAC6, expression of OsNAC5, OsNAC9, and OsNAC10 is activated by abiotic stress, and overexpression of these transcription factors significantly improved drought tolerance [105][106][107]. Consequently, grain yields of the OsNACs-overexpressing transgenic rice plants were higher than that of wild-type control plants under stress conditions. Unlike these NACs, OsNAC2 negatively regulates plant tolerance to drought stress [114], suggesting that stress-responsive transcription factors including NACs are involved in negative regulation of plant tolerance to abiotic stresses [115]. Previous studies also show that NACs mediate modulation of plant growth and development under stress conditions. For example, overexpression of OsNAC5, OsNAC9, and OsNAC10 promoted root growth, suggesting that these stress-responsive NACs are involved in developmental changes under stress conditions [116]. A wheat NAC transcription factor, TaRNAC1, showed a similar function to the rice NACs; overexpression of TaRNAC1 resulted in enhanced dehydration tolerance, and promoted root growth in wheat [117].
Despite the crucial role of NAC transcription factors in abiotic stress tolerance, the regulatory interactions between NACs and ABA, or NACs and JA remain largely unknown. However, increasing evidence suggests that NACs cooperate with ABA and JA to modulate plant responses and tolerance to abiotic stress. For example, the expression of many NACs is regulated by ABA and JA, as well as abiotic stress [103,[105][106][107]118]. Furthermore, the Arabidopsis NAC transcription factor ANAC096 directly interacts with ABF2 and 4, which are key transcription factors of ABA signaling, to regulate dehydration and osmotic stress responses, and the Arabidopsis NAC transcription factor VASCULAR-RELATED NAC-DOMAIN-INTERACTING 2 (VNI2) integrates ABA signals into leaf senescence [119,120]. In addition, ANAC019 and ANAC055 regulate the JA-signaled defense response, and the rice NAC transcription factor RIM1 functions as a regulator of JA signaling [121,122]. These findings suggest that NACs are fundamentally involved in ABA-and JA-mediated modulation of abiotic stress responses in plants.

AP2/ERF Transcription Factors
The APETALA 2/ETHYLENE-RESPONSIVE FACTOR (AP2/ERF) transcription factor family comprises a large group of plant-specific transcription factors (122 AP2/ERFs in Arabidopsis and 139 in rice) that includes four major subfamilies, AP2, RELATED TO ABSCISIC ACID INSENSITIVE 3/VIVIPAROUS 1 (RAV), ETHYLENE-RESPONSIVE FACTOR (ERF), and DEHYDRATION-RESPONSIVE ELEMENT-BINDING PROTEIN (DREB) [123][124][125]. Genome-wide transcription profiling studies showed that expression of most AP2/ERFs is activated by abiotic stress, and this activation largely depends on the stress or ABA-responsive cis-elements of their promoter regions. For example, the Arabidopsis AP2/ERF transcription factor DREB2A plays a key role in controlling plant tolerance to abiotic stresses such as drought, and the expression of DREB2A is controlled by a direct interaction between stress-signaling proteins, such as ABFs and the ABA-Responsive Element (ABRE) of the DREB2A promoter [126,127]. AP2/ERF transcription factors contain an AP2 domain consisting of approximately 60 amino acids in the N-terminal region and a regulatory domain in the C-terminal region, which are responsible for the activity of DNA binding and transcriptional regulation, respectively [125].
AP2/ERFs play an essential role in plant response and tolerance to abiotic stress by regulating the expression of stress-responsive genes. DREB1s/C-REPEAT-BINDING FACTORS (CBFs), belonging to the AP2/ERF transcription factor family, are key regulators of plant response to cold temperature stress [128]. Expression of DREB1s/CBFs is strongly upregulated by cold, and overexpression of DREB1s enhances plant tolerance to freezing stress in Arabidopsis. In contrast to the overexpressing plants, knock-out mutants that lack the activity of DREB1s displayed reduced tolerance to freezing temperatures, indicating a pivotal role of DREB1s in cold stress tolerance [129]. Expression of DREB1s also affects plant tolerance to drought and salt stress, and it is likely that expression of a common set of stress-responsive gene is involved in this phenomenon [125,130,131]. The crucial role of AP2/ERFs in plant stress tolerance was also shown in other subfamilies of the AP2/ERFs in Arabidopsis. AtERF53 is a stress-responsive AP2/ERF transcription factor and its overexpression confers enhanced tolerance to heat stress and drought in Arabidopsis [132,133]. Similar to AtERF53, transgenic plants overexpressing AtERF74 displayed enhanced tolerance to drought, whereas AtERF74 knock-out mutants exhibited reduced tolerance [133]. Furthermore, the peanut (Arachis hypogaea) AP2/ERF transcription factor AhDREB1 improved tolerance to osmotic stress in Arabidopsis [134]. These studies indicate a crucial role of AP2/ERFs in stress tolerance, and that the function of stress-responsive AP2/ERFs is conserved in plants.
Because of the essential role of AP2/ERFs in stress tolerance, many studies have aimed to identify and characterize stress-responsive AP2/ERFs in crops. For example, the rice stress-responsive AP2/ERFs OsDREB1, OsEREBP1, and OsERF71 mediate abiotic stress responses in rice. The expression of OsDREB1, OsEREBP1, and OsERF71 is regulated by abiotic stress, and their overexpression improved abiotic stress tolerance in rice by regulating the expression of stress-responsive genes [135][136][137][138][139]. Interestingly, the stress-responsive OsERF71 also regulates lignin biosynthesis by directly regulating the expression of lignin biosynthesis genes such as CINNAMOYL-COENZYME A REDUCTASE 1 [137]. Lignin biosynthesis affects drought stress because lignin, a key component of plant secondary cell walls, is hydrophobic and inhibits water loss from plant tissues [140]. Indeed, drought-tolerant inbred lines displayed increased levels of lignin compared to drought-sensitive lines in maize [141]. These studies suggest that OsERF71 regulates plant tolerance to drought stress by controlling lignin biosynthesis. Many stress-responsive AP2/ERFs have been identified in other crops besides rice, including soybean ERF3, maize DREB2A, and tomato SlERF5, based on their stress-responsive expression patterns and functions in abiotic stress tolerance [17,[142][143][144]. This further suggests that AP2/ERF transcription factors are critical for meditating plant response and tolerance to abiotic stress.
It has been suggested that DREBs regulate plant response to abiotic stress in an ABA-independent manner [145]. However, increasing evidence suggests that several stress-responsive AP2/ERFs are involved in ABA-dependent stress responses [17,132,134,137,139,146]. Furthermore, a study by Feng et al. (2014) showed that the AP2/ERF transcription factor RAV1 directly interacts with SnRK2s, which are key kinases that determine the activity of ABFs, and regulates ABA sensitivity [147]. This finding supports the idea that some AP2/ERF transcription factors are involved in ABA-dependent stress responses. In addition, a recent study showed that OCTADECANOID-RESPONSIVE AP2/ERF TRANSCRIPTION FACTOR 47 (ORA47) regulates the biosynthesis of JA, suggesting that AP2/ERFs might be involved in modulation of JA -mediated plant responses [17].

MYB Transcription Factors
MYB transcription factors are named for the conserved MYB domain responsible for DNA binding, and constitute a large family of transcription factors [148]. The MYB domain is composed of approximately 50 amino acids and MYB transcription factors are divided into four subfamilies based on the number and the position of MYB repeats: 1R-, R2R3, R1R2R3-, and 4R-MYB [149]. R2R3-MYBs are plant-specific MYB transcription factors, and it has been suggested that plants have approximately 100 R2R3-MYBs [150,151]. Genome-wide transcription profiling showed that expression of many MYBs and genes with MYB-binding elements are regulated by abiotic stresses such as drought [152][153][154][155], suggesting that MYBs are involved in plant responses and tolerance to abiotic stress. MYBs tend to interact with other transcription factors, and direct interaction between MYBs and stress-responsive transcription factors, such as MYCs and WRKYs, support the finding that MYBs are involved in plant response and tolerance to abiotic stresses [156][157][158]. The role of MYBs in abiotic stress response have been further demonstrated by functional characterization studies using overexpression and knock-out systems. For example, AtMYB44 regulates drought tolerance in Arabidopsis [159]. AtMYB44 is specifically expressed in guard cells, and the expression is activated by abiotic stresses such as drought, cold, and salinity. Furthermore, overexpression of Arabidopsis AtMYB44 increases ABA sensitivity and ABA-induced stomatal closure, leading to improved tolerance to drought stress. By contrast, atmyb44 knockout plants displayed increased sensitivity to drought stress, indicating that AtMYB44 positively regulates drought tolerance by modulating ABA-dependent stomatal closure [159]. AtMYB96 is also involved in drought tolerance in Arabidopsis [160,161]. Expression of AtMYB96 is induced by drought and ABA, and overexpression of AtMYB96 enhanced drought resistance by activating cuticular wax biosynthesis, which inhibited water loss through the leaf surface. This result is supported by a study in Camelina sativa in which heterologous overexpression of AtMYB96 resulted in improved drought tolerance via cuticular wax accumulation, similar to Arabidopsis [162], suggesting that the function of AtMYB96 in cuticular wax synthesis and drought tolerance is conserved among plants.
MYBs are also involved in salt stress. Expression of AtMYB20 is upregulated by salt stress, and transgenic plants overexpressing AtMYB20 showed improved tolerance to salt stress [163]. By contrast, suppression of AtMYB20 resulted in hypersensitivity to salt stress, indicating that AtMYB20 positively regulates plant tolerance to salt stress. Similar to Arabidopsis, many stress-responsive MYBs have been identified in crops, including rice and soybean, based on their expression patterns and functions in abiotic stress tolerance. For example, OsMYB4, OsMYB6, OsMYB48-1, and OsMYB91 were identified in rice, and GmMYB76, GmMYB92, and GmMYB177 in soybean [164][165][166][167][168]. The interaction between MYBs and ABA is largely unknown, but several studies showing that the expression of many stress-responsive MYBs is also regulated by ABA and JA suggest that MYBs are involved in modulation of ABA-and JA-dependent stress responses. This is partially supported by studies showing that AtMYB20 suppresses the expression of PP2Cs, which encode negative regulators of ABA signaling, and AtMYB21 and 24 directly interact with the JA signaling repressors JAZs [169].

WRKY Transcription Factors
WRKY transcription factors contain a highly conserved WRKYGQK motif in their N-terminal region and a novel zinc-finger-like motif at their C-terminal region [170]. WRKY transcription factors constitute a large family in plants. For example, the Arabidopsis, rice, and soybean genomes are predicted to have approximately 74, 103, and 197 WRKYs, respectively [170][171][172]. Stress-responsive expression patterns of WRKYs have suggested their involvement in plant response to abiotic stress. For example, a microarray analysis showed that the expression of 18 WRKYs (among 35 WRKYs tested) is induced by salt stress [100]. The involvement of WRKYs in plant stress responses has been demonstrated through functional characterization using knock-out or overexpression systems. In Arabidopsis, heat stress affects the expression of AtWRKY25, 26, and 33 [173]. In normal growth conditions, germination and growth of wrky25 wrky26 and wrky25 wrky33 double mutants and the wrky25 wrky26 wrky33 triple mutant were similar to those of wild-type control plants. However, these mutants were hypersensitive to heat stress, and their survival rate was significantly decreased. By contrast, transgenic plants overexpressing WRKY25, WRKY26, and WRKY33 displayed enhanced tolerance to heat stress, providing evidence that WRKY25, WRKY26, and WRKY33 positively regulate plant tolerance to heat stress. Similarly, heat stress-responsive WRKY39 also positively regulates heat stress tolerance [174]. In addition, increasing evidence suggests that WRKYs are involved in modulating development under stress conditions, similar to NACs and AP2/ERFs. For example, stress-responsive WRKY46 modulates the development of lateral roots under abiotic stress conditions, and WRKY22 mediates dark-induced senescence [175,176]. Despite the crucial role of WRKYs, the interaction between WRKYs and ABA in abiotic stress response remains largely unknown; however, findings that the expression of WRKYs, such as AtWRKY18, AtWRKY40, AtWRKY60, AtWRKY63, and GhWRKY17, affects the ABA response suggest that WRKYs are involved in ABA-mediated abiotic stress signaling [177][178][179].
Several WRKYs have been identified in crops and functionally characterized through gain-of-function studies. Overexpression of the stress-induced rice OsWRKY11 and OsWRKY45 enhanced drought tolerance, and heterologous overexpression of wheat TaWRKY1 and TaWRKY33 also improved drought tolerance in Arabidopsis [180][181][182]. These studies suggest that the function of stress-responsive WRKYs is conserved in plants, and also propose that stress-responsive WRKYs may be promising target genes for improving abiotic stress tolerance in crops. The transcription factors described in this review were summarized in Table 1.

Future Perspectives
ABA and JA are key hormones that mediate plant response and tolerance to abiotic stress. However, increasing evidence suggests that other phytohormones, such as auxin, cytokinin, brassinosteroid, salicylic acid and ethylene, are also involved in plant response to abiotic stress [183][184][185]. At present, a large number of phytohormone-derived growth regulators are used commercially in agriculture to improve plant tolerance to abiotic stresses [186][187][188][189][190][191][192][193], suggesting that identification and characterization of genetic components such as transcription factors, which mediate hormonal response, provide important clues to understand the molecular mechanisms underlying plant response and tolerance to abiotic stress, and to develop crops with improved stress tolerance. In this review, we summarized the function of stress-responsive transcription factors in modulating plant response and tolerance to abiotic stress, focusing on bZIPs, bHLHs, NACs, AP2/ERFs, MYBs, and WRKYs. However, other families of transcription factors, such as HD-ZIP and Zinc Finger transcription factors, whose expressions are regulated by abiotic stress, are also involved in plant response and tolerance to abiotic stress [194][195][196][197]. Although many of the questions about the molecular mechanisms underlying transcriptional and translational interaction between multiple transcription factors remain largely unanswered, these suggest that modulation or manipulation of transcription involved in plant stress response factors would be a good strategy to enhance tolerance to abiotic stress in crops. Numerous studies have identified and characterized the roles of stress-responsive transcription factors in plant response and tolerance to abiotic stress, but most of these studies were performed in laboratory conditions. Functional validation under field conditions will expand potential applications of stress-responsive transcription factors for improvement of abiotic stress tolerance in crops.