The guard cell turgor is dynamically adjusted to environmental conditions and hormonal signals in order to facilitate the proper gas exchange and prevent excessive water loss. Mature guard cells do not have plasmodesmata and for this reason most influx and efflux of solutes occurs via ion channels, transporters, and pumps that are localized in the plasma membrane (PM). The action of ion channels, transporters, and pumps that are essential for stomatal function is well documented and supported by molecular studies involving mutants in the genes encoding these protein. During the opening of the stomata, the H⁺-ATPase pump mediates the efflux of H⁺ from the guard cells. In plants, H⁺-ATPases belong to the multi-gene family of the P-type ATPases, with 11 genes in Arabidopsis, which are all expressed in the guard cells (Ueno et al., 2005). In the guard cells, the action of H⁺-ATPase activity is positively regulated by blue light and auxins, whereas Ca²⁺ and ABA act as negative regulators. The efflux of H⁺ hyperpolarizes the PM and leads to K⁺ uptake via activation of inward K⁺ rectifying channels, such as KAT1 (potassium channel in Arabidopsis thaliana 1), KAT2 (potassium channel in Arabidopsis thaliana 2), and AKT1 (Arabidopsis thaliana K⁺ transporter 1) (Schachtman et al., 1992; Pilot et al., 2001; Szyroki et al., 2001). Another signal that activates the influx of K⁺ via K⁺ channels is the acidification of the apoplast as a result of H⁺ extrusion from the guard cells. K⁺ uptake is balanced by counter-ions, mainly Cl⁻ obtained from the apoplast, malate²⁻ that is derived from starch breakdown or NO3-. The last one is transported from the apoplast by a nitrate transporter AtNRT1.1 (CHL1) [nitrate transporter 1 (chlorina1)]. The importance of NO3- uptake was confirmed by an analysis of an Arabidopsis clh1 mutant. The stomatal apertures of the chl1 mutant were smaller than those of the wild-type when nitrate was supplied. Furthermore, the chl1 mutant was drought tolerant (Guo et al., 2003). Ions supplied into the guard cells together with water transported via aquaporins generate the turgor that are necessary to keep stomata open (Figure 1A).

During stomatal closure, the inhibition of H⁺-ATPase and the activation of anion channels together result in membrane depolarization. Anion channels such as rapid channels (R-type) and slow channels (S-type) facilitate the efflux of malate²⁻, Cl⁻, and NO3- (Roelfsema et al., 2004; Roelfsema and Hedrich, 2005). The decreased level of malate²⁻ in guard cells is also linked with the gluconeogenic conversion of malate²⁻ into starch (Willmer and Fricker, 1996). Membrane depolarization creates a driving force for the efflux of K⁺ via K⁺ outwardly rectifying channels such as GORK (guard cell outwardly rectifying K⁺ channel) (Jeanguenin et al., 2008). An Arabidopsis gork mutant displayed impaired stomatal closure, thus confirming the important role of GORK in elimination K⁺ ions and in the facilitation of stomatal closure (Hosy et al., 2003). Another event that accompanies stomatal closure is an elevation of the cytoplasmic Ca²⁺ concentration as a result of Ca²⁺-release via channels situated in both the PM and in the tonoplast (MacRobbie, 2006). Ca²⁺ channels are encoded by genes from three gene-families: TPC1 (two-pore channel 1) (Peiter et al., 2005), CNGC (cyclic nucleotide gated channel) (Finn et al., 1996), and GLR (glutamate receptor) (Lacombe et al., 2001). Taken together, the efflux of solutes from the guard cells leads to a reduced turgor and stomatal closure (Figure 1B).

Abscisic acid has been postulated as a main regulator of stomatal movements but its proper functioning depends on the appropriate level of biologically active ABA within the plant cells. This is achieved by synchronized processes such as ABA biosynthesis, catabolism, conjugation/deconjugation, and transport. These processes, which are well recognized and studied in various species, have confirmed the function of many enzymes involved in the biosynthesis, catabolism, conjugation/deconjugation, and transport of ABA. The exception, not fully recognized yet, is ABA signal transduction pathway. Although ABA has been the focus of many research groups since the early 90s, there are still many questions in regards to the function of the proteins involved in ABA signaling, protein interactions or the impact of the components of signalosome on specific physiological responses. Therefore, with the progress in studies on ABA signaling, the state of knowledge and the already known interaction web should be updated and verified.

Abscisic acid is synthesized in the plastids and cytosol, mainly in the vascular parenchyma cells but also in the guard cells, through the cleavage of a C40 carotenoid precursor, followed by a two-step conversion of the intermediate xanthoxin into ABA via ABA-aldehyde (Taylor et al., 2000; Finkelstein and Rock, 2002; Schwartz et al., 2003; Endo et al., 2008; Melhorn et al., 2008). The pathway begins with isopentenyl pyrophosphate (IPP), which is the biological isoprene unit and the precursor of all terpenoids, as well as many plant hormones. The next step is the epoxidation of zeaxanthin and antheraxanthin into violaxanthin, which is then catalyzed by zeaxanthin epoxidase (ZEP) (Marin et al., 1996). After a series of violaxanthin modifications that are controlled by the enzyme ABA4, violaxanthin is converted into 9-cis-epoxycarotenoid (North et al., 2007). Oxidative cleavage of the major epoxycarotenoid 9-cis-neoxanthin by the 9-cis-epoxycarotenoid dioxygenase (NCED) yields a C15 intermediate – xanthoxin (Schwartz et al., 1997). This step is the last one that occurs in plastids. Xanthoxin is exported to the cytoplasm where a two-step reaction via ABA-aldehyde occurs. The first step is catalyzed by a short-chain alcohol dehydrogenase/reductase (SDR) that is encoded by the AtABA2 (ABA deficient 2) gene (Rook et al., 2001; Cheng et al., 2002; Gonzalez-Guzman et al., 2002) and that generates ABA-aldehyde. Then, the ABA-aldehyde oxidase (AAO) with the molybdenum cofactor (MoCo) catalyzes the last step in the biosynthesis pathway – the conversion of ABA-aldehyde into ABA (Seo et al., 2004) (Figure 2A). The appropriate level of active ABA is achieved not only through the biosynthesis and catabolism reactions performed by CYP707A1-4 (cytochrome P450, family 707, subfamily A, polypeptide 1, 2, 3, 4) (Kushiro et al., 2004; Figure 2B), but also by the inactivation of ABA through conjugation and deconjugation. ABA can be inactivated at the C-1 hydroxyl group by different chemical compounds that form various conjugates and that accumulate in vacuoles or in the apoplastic space (Dietz et al., 2000). The most widespread conjugate is ABA glucosyl ester (ABA-GE), which is catalyzed by ABA glucosyltransferase (Boyer and Zeevaart, 1982). Lee et al. (2006) identified the AtBG1 (beta-1,3-glucanase 1) protein that is responsible for the release of ABA from ABA-GE. Their findings showed that ABA deconjugation plays a significant role in providing an ABA pool that allows plants to adjust to changing physiological and environmental conditions (Figure 2C).

The ability of ABA to move long distances allows it to serve as a critical stress messenger. Kuromori et al. (2011) identified the ABA importer – ABCG22 (Arabidopsis thaliana ATP-binding cassette G22). The gene encoding this transporter is mainly expressed in the guard cells. In addition, the expulsion of ABA into the intercellular space is mediated by transporters such as ABCG25 (Arabidopsis thaliana ATP-binding cassette G25). ABCG25 is expressed primarily in vascular tissues where ABA is synthesized (Kuromori et al., 2010). ABA delivery to the guard cells promotes a cascade of reactions that lead to stomatal closure and that inhibit stomatal opening in order to prevent water loss (Figure 2D).

After ABA is received from ABC transporters by the guard cells, the PYR/PYL/RCAR (pyrabactin-resistance 1/pyrabactin-resistance like/regulatory component of ABA receptor) perceives ABA intracellularly and forms complexes that inhibit clade A of PP2Cs (protein phosphatase 2C), the negative regulators of ABA signaling, such as ABI1 (ABA insensitive 1), ABI2 (ABA insensitive 2), HAB1 (hypersensitive to ABA1) (Ma et al., 2009; Park et al., 2009; Santiago et al., 2009; Nishimura et al., 2010). The inactivation of PP2Cs allows downstream targets to be phosphorylated and activated – Sucrose Non-fermenting 1-Related subfamily 2 protein Kinases (SnRK2) (Fujii and Zhu, 2009; Fujita et al., 2009; Umezawa et al., 2009; Kim et al., 2010). ABA receptors, PP2Cs, and SnRKs form the core of the early ABA signaling cascade (Figure 2E).

The ABA mode of action is linked to diurnal stomatal movements. It has been proposed that this link is based on both the molecular connections between ABA and circadian-clock pathways and on ABA biosynthesis and response to light (reviewed in Tallman, 2004). Although several studies have been carried out linking the diurnal cycle with ABA signaling, there is still a need for further research that would clarify this connection. It has been confirmed that the elevated ABA levels in the dark phase of the day are responsible for stomatal closure but, on the other hand, the molecular basis of the sensing CO₂ molecules by guard cells is still not well understood. This part of investigations still needs confirmation through the use of well-established methods.

In darkness, stomata are closed. This is probably caused by an intensive ABA accumulation through the biosynthesis of ABA in the guard cells and the simultaneous import of endogenous ABA from the apoplast to the guard cells using ABA transporters such as ABCG22 (Kuromori et al., 2011), while at the same time, ABA catabolism processes are disfavored. Elevated ABA levels cause stomata closure via the activation of an ABA signaling cascade, the efflux of Ca²⁺ from internal stores, the activation of S-type and R-type anion channels that lead to the efflux of Cl⁻, malate²⁻, and NO3- and the activation of the GORK channel that leads to the efflux of K⁺. During the night, elevated levels of CO₂ in the leaves were observed due to respiration. It has been proved that CO₂ has a positive effect on the stomatal closure process. The guard cells probably do not sense CO₂ molecules but instead HCO3- is synthesized from CO₂ (Hu et al., 2010), which activates S-type channels and leads to the efflux of Cl⁻, malate²⁻, and NO3- (Xue et al., 2011) (Figure 3A).

At first light, a depletion of endogenous ABA is observed through xanthophyll cycling, the isomerization of ABA precursors and the activation of ABA catabolism enzymes, such as CYP450 (cytochrome P450). The degradation of ABA liberates the guard cells to extrude H⁺ via H⁺-ATPase (H⁺-pump) and accumulate water and ions, such as K⁺, Cl⁻, malate²⁻ in order to generate the turgor needed to keep stomata open. K⁺ uptake is mainly responsible for the rapid increase of the turgor and the opening of stomata during the dawn (Humble and Raschke, 1971; Talbott and Zeiger, 1996). The accumulation of sugars such as glucose, fructose and sucrose has been reported during the light phase of the day (Talbott and Zeiger, 1998). In the midday, ABA is delivered to the apoplast around the guard cells through the xylem transpiration stream and the guard cells are regulated by steady-state ABA concentrations (Figure 3B).

In the evening, ABA biosynthesis outweighs the ABA catabolism in the guard cells, which leads to stomatal closure (for review, see Tallman, 2004).

Under drought stress conditions, ABA would reach a concentration high enough to cause ion efflux and an inhibition of sugar uptake by the guard cells in the midday, thus reducing the apertures for the rest of the day. Analyses of ABA biosynthesis, catabolism, de/conjugation, and transport have been supported by various studies involving several species and different methods, such as mutant analysis, transcriptomics, proteomics, or immunohistochemical techniques. In order to define the role of ABA in stress response, the action of several components of the pathways mentioned were tested in response to stress. The engagement of such various techniques makes the state of knowledge in the field of ABA biosynthesis, catabolism, de/conjugation, and transport well supported and reliable.

It has been shown that ABA concentrations can increase up to 30-fold in response to drought stress (Outlaw, 2003). Water deficit promotes ABA biosynthesis via the upregulation of a key enzyme – NCED3. A significant increase in NCED transcript levels can be detected within 15–30 min after leaf detachment or dehydration treatment (Qin and Zeevaart, 1999; Thompson et al., 2000), which indicates that the activation of NCED genes can be fairly quick. Cheng et al. (2002) reported that the AtNCED3, AtZEP (Zeaxanthin epoxidase), and AtAAO3 (ABA-aldehyde oxidase) genes could be induced in Arabidopsis by ABA and studies in rice showed that OsNCED3 expression was induced by dehydration (Ye et al., 2011). An immunohistochemical analysis, using antibodies raised against AtNCED3, revealed that protein is accumulated in the leaf vascular parenchyma cells in response to drought stress. This was not detected in non-stressed conditions. These data indicate that drought-induced ABA biosynthesis occurs primarily in the vascular parenchyma cells and that vascular-derived ABA might trigger stomatal closure via the transport to the guard cells (Endo et al., 2008). AtNCED3 expression is upregulated by drought conditions across the species observed and decreases after rehydration.

Drought, like the dark part of a diurnal cycle, also promotes the deconjugation of the ABA-glucose ester (ABA-GE), which is stored in the vacuoles of leaf cells and also circulates in the plant (Xu et al., 2002; Seiler et al., 2011). Both processes, intensive ABA biosynthesis and ABA deconjugation, lead to the accumulation of high levels of biologically active ABA. ABA delivery to the guard cells via ABCG transporters, such as AGCG22 that was mentioned above, promotes a cascade of reactions that lead to stomatal closure and that inhibit stomatal opening in order to prevent water loss (Figure 2).

The ABA signaling network that leads to stomatal closure under stress is activated by the perception ABA. This begins a cascade of reactions that leads to the reduced turgor of the guard cells through ABA modulation of ion channel activities, including the regulated efflux of anions and potassium ions and the inhibition of K⁺ import. Recently, the core signalosome of ABA signaling including ABA receptors, phosphatases (PP2Cs), and kinases (SnRK2s) was established (Ma et al., 2009; Park et al., 2009; Santiago et al., 2009; Nishimura et al., 2010). Although its function is clear and confirmed by advanced molecular analysis, there is still a need to explain the impact of single components, such as kinases, on the regulation of the ion channels or the proton pump (e.g., AHA1), which is described below. On the other hand, the interaction between ABA regulated kinases SnRK2s and S-type anion channels, and the potassium inwardly rectifying channels, described below, has been well established and documented.

The inactivation of PP2Cs such as ABI1 and ABI2 by the complex ABA-receptor facilitates the phosphorylation and activation of a downstream target of phosphatases – SnRK2, such as SnRK2.2/D, SnRK2.3/E, and SnRK2.6/OST1/E, which are the key players in the regulation of ABA signaling and abiotic stress response (Fujii and Zhu, 2009; Fujita et al., 2009; Umezawa et al., 2009). Kinases are able to regulate the activity of ion channels and the proton pump. It was shown that ABA inhibits the action of a proton pump such as H⁺-ATPase. The dominant Arabidopsis mutant ost2 (opened stomata 2) in AHA1 (H⁺-ATPase 1 HA1) gene exhibited the constitutive activation of AHA1 H⁺-ATPase, which in turn resulted in an inability to close stomata in response to ABA (Merlot et al., 2007). The molecular mechanism of the inhibition of AHA1 by ABA has not yet been fully elucidated. One of the most direct pieces of evidence of the regulation of H⁺-ATPase by SnRK is the demonstration that specific calcium-stimulated kinase, PSK5 (a member of the SnRK3 kinase family), is able to phosphorylate the closest homolog of AHA1 – AHA2 (H⁺-ATPase 1 HA2) in Ser392 localized in the C-terminus of the AHA2 protein. This reaction prevents the 14-3-3 protein, which is the main activator of AHA2 leading to the inhibition of H+-ATPase action, from binding (Fuglsang et al., 2007).

Sucrose Non-fermenting 1-Related subfamily 2 protein Kinases also regulate S-type anion channels and potassium inwardly rectifying channels such as SLAC1 (slow anion channel-associated 1) and KAT1 (K⁺ channel in Arabidopsis thaliana), respectively. The first one is activated by SnRK2, whereas KAT1 is inhibited. SLAC1 encodes the anion-conducting subunit of an S-type anion channel. In different species, S-type anion channels are activated in the guard cells by ABA, cytosolic Ca²⁺, and phosphorylation events (Schmidt et al., 1995; Pei et al., 1997; Leonhardt et al., 1999; Raschke et al., 2003; Roelfsema et al., 2004; Mori et al., 2006). The slac1 mutant displayed a strongly impaired response to a range of stomatal closing stimuli such as ABA and Ca²⁺ (Negi et al., 2008; Vahisalu et al., 2008). Increased SLAC1 activity causes an efflux of anions which results in depolarization of the membrane as a consequence of phosphorylation by SnRK. This in turn leads to the loss of K⁺ cations from the cell through the K⁺ efflux channel GORK (guard cell outward-rectifying K⁺), which is activated by depolarization (Jeanguenin et al., 2008). KAT1 is an inward K⁺ channel that allows an influx of K⁺ inside the guard cell when the proton pump drives the PM to a negative potential. When plants encounter drought stress conditions and the ABA level rises, both the proton pumps (as mentioned above) and KAT1 are inactivated by SnRKs. It was shown that the activity of KAT1 is inhibited by an elevation of ABA and cytosolic Ca²⁺ (Schroeder and Hagiwara, 1989; Blatt and Armstrong, 1993; Grabov and Blatt, 1999) via phosphorylation by SnRK, which in turn results in a decreased influx of K⁺ into the guard cells (Hubbard et al., 2010). The loss of K⁺ and anions from the guard cells is accompanied by the efflux of water via aquaporins. Together, these events lead to a reduction of the turgor, which results in stomatal closure in response to ABA as a major signal of drought (Figure 3A).

Abscisic acid activates the Ca²⁺-permeable channels in the PM of the guard cells and triggers an influx of Ca²⁺ into the cytoplasm of the guard cells through the release of the second messenger, inositol-1,4,5-triphosphate (IP₃), which in turn activates the Ca²⁺ channels that are located in the vacuole and endoplasmic reticulum (Schroeder and Hagiwara, 1990; Hamilton et al., 2000; Krinke et al., 2007; Kwak et al., 2008). Ca²⁺-dependent protein kinases (CDPKs) are activated during drought stress and are able to control stomatal closure in an ABA-dependent manner. After ABA is perceived by a receptor, the action of PP2Cs such as ABI1 are inhibited. ABI1 was identified as a negative regulator of CPK21 (Ca²⁺ dependent protein kinase 21), which like SnRK phosphorylates SLAC1. SLAC1 phosphorylation, in turn, results in the activation of anion and the efflux of K⁺ (Geiger et al., 2010). An increased cytosolic Ca²⁺ level activates the Ca²⁺-dependent pathways that inhibit K⁺ import and trigger the depolarization of the membrane. Mori et al. (2006) identified two calcium-dependent kinases – CPK3 (calcium-dependent protein kinase 3) and CPK6 (calcium-dependent protein kinase 6) as positive regulators of ABA signaling in the guard cells during water stress. Inactivation of both genes led to a reduction in the activation of S-type channels by ABA and Ca²⁺, the impairment of the ABA activation of Ca²⁺ permeable channels and a decreased sensitivity of stomata to ABA. Disruption of the regulatory subunit RCN1 (roots curl in NPA) of the gene encoding PP2A (protein phosphatase 2A) led to a reduction of the ABA activation of anion channels and a decreased sensitivity of stomata to ABA (Kwak et al., 2002, Figure 4).

Another protein involved in ABA signaling in stomata is GPA1. GPA1 is a positive regulator in the ABA-mediated inhibition of stomatal opening. Arabidopsis plants lacking GPA1 (Gα subunit of G protein) showed a reduction in the inhibition of inward K⁺ currents and a reduced guard cell ABA-insensitivity in stomatal opening (Wang et al., 2001). The mutants era1 (enhanced response to ABA1) and abh1 (ABA hypersensitive 1), which are deficient in a farnesyl transferase subunit and RNA cap-binding protein, respectively, are ABA hypersensitive and showed an enhanced ABA activation of S-type channels (Pei et al., 1998; Schroeder et al., 2001a; Hugouvieux et al., 2002; Figure 4). However, the exact molecular role of ERA1 or ABH1 in stomatal regulation should be clarified in future research.

During stomatal closure, slow vacuolar (SV) channels activated by cytosolic Ca²⁺ contribute to the release of Ca²⁺ from vacuoles. SV channels were shown to be calcium permeable and it was suggested that they facilitate a brief transient efflux of cations, including Ca²⁺, from vacuoles (Ward and Schroeder, 1994).

Several of the genes involved in the processes described above and more are presented in Table 1 together with a description of mutant phenotypes.

The guard cells generate reactive oxygen species (ROS) such as hydrogen peroxide (H₂O₂) and nitric oxide (NO) in response to ABA (Pei et al., 2000; Zhang et al., 2001). Exogenous H₂O₂ activates permeable Ca²⁺ channels in the PM of Arabidopsis guard cells and inhibits inward K⁺ channels (Zhang et al., 2001; Kohler et al., 2003; Kwak et al., 2003). Taking into account the fact that ROS and NO signaling is not yet fully understood, there is a need for further analysis in order to elucidate their function, for example, the role of Ca²⁺ in ROS and NO action in guard cells should be clarified.

Reactive oxygen species production in Arabidopsis guard cells is mediated by two subunits of NADPH oxidase – AtrbohD (Arabidopsis thaliana respiratory burst oxidase homolog D) and AtrbohF (Arabidopsis thaliana respiratory burst oxidase homolog F). The significance of ROS involvement in stomatal closure was revealed by an analysis of the atrbohD/atrbohF double mutant, which showed impaired stomatal closure in response to ABA (Kwak et al., 2003). The protein, OST1 (open stomata1), displays dominant kinase activity during drought stress response and is able to activate NADPH oxidase (Sirichandra et al., 2009). Mutants in OST1 showed a wilty phenotype in water deficit conditions because of the impairment of stomatal closure and ROS production (Mustilli et al., 2002; Yoshida et al., 2006; Figure 4).

Another crucial factor for stomatal closure is NO, which is generated in response to ABA (Neill et al., 2002, 2008). Exogenously applied NO donors triggered stomatal closure, whereas the application of an NO scavenger inhibited ABA-induced stomatal closure (Neill et al., 2002; Figure 4).

There is some evidence that both H₂O₂ and NO actions in the guard cells require calcium. In addition, H₂O₂ inhibits K⁺ channel activity, induces cytosolic alkalization in the guard cells and promotes NO signaling in response to ABA (Zhang et al., 2001; Kohler et al., 2003; Wang and Song, 2008). Conversely, NO neither stimulates H₂O₂ synthesis nor does it require H₂O₂ for its action (Bright et al., 2006).