PYR/RCAR Receptors Contribute to Ozone-, Reduced Air Humidity-, Darkness- and CO 2 -Induced Stomatal Regulation

One-sentence summary: Signaling through abscisic acid PYR/RCAR receptors plays a fundamental role in controlling whole-plant stomatal conductance and affects plant stomatal closure in response to low air humidity, darkness, O 3 pulse and elevated CO 2 . ABSTRACT Rapid stomatal closure induced by changes in the environment, such as elevation of CO 2 , reduction of air humidity, darkness and pulses of air pollutant ozone (O 3 ), involves the SLAC1 anion channel. SLAC1 is activated by OST1 and Ca 2+ -dependent protein kinases. OST1 activation is controlled through abscisic acid (ABA)-induced inhibition of PP2C protein phosphatases by PYR/RCAR receptor proteins. To address the role of signaling through PYR/RCARs for whole-plant steady-state stomatal conductance (g st ) and stomatal closure induced by environmental factors, we used a set of Arabidopsis thaliana mutants defective in ABA metabolism/signaling. The g st values varied several fold among the studied mutants indicating that basal ABA signaling through PYR/RCAR receptors plays a fundamental role in controlling whole-plant water loss through stomata. PYR/RCAR-dependent inhibition of PP2Cs was clearly required for rapid stomatal regulation in response to darkness, reduced air humidity and O 3 . Furthermore, PYR/RCAR proteins seem to function in a dose-dependent manner and there is a functional diversity among them. Although rapid stomatal response to elevated CO 2 was evident in all but slac1 and ost1 mutants, the bicarbonate-induced activation of S-type anion channels was also reduced in dominant active PP2C mutants, abi1-1 and abi2-1 . Further experiments with wider range of CO 2 concentrations and analyses of stomatal response kinetics suggested that ABA signalosome affects the CO 2 –induced stomatal response. Thus, we show that PYR/RCAR receptors play an important role for the whole-plant stomatal adjustments and responses to low humidity, darkness and O 3 and are involved in responses to elevated CO 2 . presented in this study suggest that CO 2 stomatal acid 3- ). Results presented in study suggest that CO 2 -induced stomatal closure and bicarbonate-induced activation of S-type anion channels is partly controlled by ABA-signalosome. Question marks highlight that the nature of signal perception at the plasma membrane and signal transduction in the cytosol leading to the activation of ABA signalosome remain to be addressed. (B) Summary of stomatal responses to closure-inducing stimuli in studied mutants. Mutants are presented in the order of phenotypic severity. O 3 , H, D and CO 2 indicate environmental factors ozone, air humidity, darkness and CO 2 , respectively. “+” indicates that either initial change in stomatal conductance or curve fitting was different from WT; “++” indicates that initial change in stomatal conductance and curve fitting were both different from WT. Absence of symbol indicates WT-like stomatal closure. Mutants that showed WT-like stomatal responses to all stimuli ( 148 , cpk21 , cyp707a1 , cyp707a3 ) are not listed in the table.

Stomata, small pores in the leaf epidermis, are formed by a pair of guard cells that have developed mechanisms to sense and respond to various endogenous and environmental stimuli. Stomata close in response to reduction in air humidity, darkness, and CO 2 enrichment. Ozone (O 3 ), a major secondary air pollutant with adverse impacts on global vegetation (Ashmore, 2005) and climate change (Sitch et al., 2007), has also been shown to cause rapid stomatal closure (Hill and Littlefield, 1969;Vahisalu et al., 2010). The key endogenous factor triggering stomatal closure in response to drought is the plant hormone abscisic acid (ABA). Webb and Hetherington (1997) suggested that the pathways of ABA-and CO 2 -induced closure converge (i.e. that there is an economy in signaling pathways leading to the promotion of stomatal closure). However, the location of this convergence point is still under debate. Opening and closure of stomatal pores is achieved by the uptake and release of osmotically active ions, leading to expanding and shrinking of guard cells. Thus, the activation and inactivation of guard cell ion channels and transporters are the primary targets of signaling networks controlling stomatal movements (for review, see Kim et al., 2010;Kollist et al., 2011;Roelfsema et al., 2012).
In 2009, independent groups simultaneously discovered the functional and structural mechanisms of ABA sensing by cytosolic PYRABACTIN RESISTANCE1 (PYR1)/PYR1-like (PYL)/REGULATORY COMPONENTS OF ABA RECEPTORS (RCAR) receptor proteins (Ma et al., 2009;Park et al., 2009). Identification of the guard cell slow-type anion channel gene, SLAC1, is another substantial finding in stomatal research (Negi et al., 2008;Vahisalu et al., 2008). ABA-induced stomatal closure involves the activation of SLAC1, which is controlled by PYR/RCAR-dependent sequestration of type 2 protein phosphatases (PP2Cs), such as ABI1, ABI2, HAB1, and PP2CA, and the concomitant activation of the Snf1-related subfamily 2 protein kinase SnRK2.6/ OST1 (Ma et al., 2009;Nishimura et al., 2010;Park et al., 2009;Umezawa et al., 2009;Vlad et al., 2009;Weiner et al., 2010;Dupeux et al., 2011;Soon et al., 2012). Phosphorylation by OST1 activates several proteins, including SLAC1 (Geiger et al., 2009;Lee et al., 2009;Vahisalu et al., 2010). SLAC1 can also be activated by different calcium-dependent protein kinases (CPKs; Geiger et al., 2010, Brandt et al., 2012Scherzer et al., 2012). Thus, several molecular details are known for the general activation mechanism of SLAC1 and subsequent stomatal closure. The importance of SLAC1 and OST1 in rapid stomatal responses to environmental factors such as darkness, CO 2 , humidity, and O 3 is also established (Negi et al., 2008;Vahisalu et al., 2008;Ache et al., 2010;Xue et al., 2011). However, whether the ABA-and PYR/ RCAR-dependent inhibition of PP2Cs that ultimately results in the activation of OST1 and anion channels like SLAC1 (herein defined as the ABA signalosome) is also required for rapid stomatal responses to important environmental factors is not fully resolved. One approach to address this question is to perform a side-byside comparison of whole-plant stomatal responses of plants where various proteins of the ABA signalosome are mutated ( Fig. 1; Table I).
Stomatal responses of several of these mutants to environmental factors have been studied previously, although often with different results. For example, initial in vitro studies indicated that OST1 was not involved in CO 2 -induced stomatal signaling, as the stomata of ost1-1 and ost1-2 behaved like the wild type in response to low CO 2 (Mustilli et al., 2002). Similarly, mutations in OST1 did not affect stomatal regulation by light, leading to the suggestion that OST1 is specifically involved in ABA signaling (Mustilli et al., 2002). However, recently, it was shown that OST1 is a major regulator of CO 2induced stomatal closing and activation of the S-type anion channels in guard cells (Xue et al., 2011). Furthermore, dominant hypermorphic abi1-1 and abi2-1 mutations, which generate mutant PP2Cs that are refractory to inhibition by PYR/RCAR receptors (Ma et al., 2009;Park et al., 2009;Umezawa et al., 2009), have been used to address whether ABA-, CO 2 -, and light-induced stomatal signaling pathways converge. Stomatal opening induced by light and CO 2 removal was 50% reduced in abi1-1 and abi2-1 plants; however, it was concluded that this might have been caused by constitutively more open stomata of these mutants (Leymarie et al., 1998a(Leymarie et al., , 1998b. Other studies showed that light-induced stomatal opening was not disrupted in abi1-1 and abi2-1 mutants (Roelfsema and Prins, 1995;Eckert and Kaldenhoff, 2000). In contrast, Webb and Hetherington (1997) found that abi1-1 and abi2-1 did not respond clearly to elevated CO 2 , whereas the stomatal closure was indistinguishable from the wild type in ABA-deficient plants.
The involvement of the ABA signalosome in the regulation of the stomatal response to reduced air humidity is also disputed. During a genetic screen for mutants involved in the stomatal response to reduced air humidity, new alleles for OST1 and ABA2, an enzyme involved in ABA biosynthesis, were identified, suggesting that OST1 activity and ABA biosynthesis are essential for low-humidity-induced stomatal closure (Xie et al., 2006). Furthermore, activation of OST1 is induced by lowhumidity stress (Yoshida et al., 2006). Contrarily, aba1, abi1-1, and abi2-1 mutants had wild-type stomatal responses to reduced humidity (Assmann et al., 2000). Since partial Figure 1. Schematic overview of ABA-induced signaling by PYR/RCAR receptors leading to sequestration of PP2Cs, activation of protein kinases OST1, CPK21, and CPK23, and subsequent phosphorylation/ activation of the SLAC1 anion channel that is essential for anion efflux and stomatal closure. Mutants selected for this study are shown. Detailed descriptions of the mutants are provided in Table I response to low humidity is observed in all studied mutants of the ABA signalosome, the existence of a separate ABAindependent pathway mediating low-humidity-induced stomatal closure has been proposed (Xie et al., 2006).
A large amount of data for stomatal signaling is collected by using isolated leaf epidermises or guard cell protoplasts, whereas the significance of these results is not always tested in intact plants. Thus, gas-exchange SALK_099139 T-DNA insertion in SLAC1 Vahisalu et al. (2008) slac1-7 (Col-0) C to T at position 527 Point mutation of Ser-120 to Phe in SLAC1 Vahisalu et al. (2010) analysis of whole-plant and leaf stomatal responses is important in verifying the stomatal responsiveness to environmental stimuli. In this study, we used a custommade gas-exchange device with parallel recording of stomatal responses of up to eight Arabidopsis (Arabidopsis thaliana) plants to various environmental factors (Kollist et al., 2007;Vahisalu et al., 2008Vahisalu et al., , 2010. We chose mutants where different components of the ABA signalosome are affected ( Fig. 1) to address the role of ABA signaling through PYR/RCAR proteins in stomatal responses to reduced air humidity, darkness, and elevated CO 2 and O 3 concentrations. We found that ABA signaling through PYR/RCARs is clearly required for rapid stomatal closure in response to darkness, O 3 , and reduced air humidity, while it is also partially involved in stomatal responses to elevated CO 2 . Since functional OST1 and SLAC1 were important in response to all stimuli, we discuss the possibility that other signaling elements besides ABA and its signalosome are able to activate OST1 in response to changes in CO 2 .

RESULTS
To study the role of the ABA signalosome in the regulation of stomatal responses to darkness, CO 2 , reduced humidity, and O 3 , we used mutants where different proteins of the ABA signalosome were affected ( Fig. 1). If possible, at least two mutants for each protein were analyzed, and in many cases, mutants from different genetic backgrounds, Columbia (Col-0) and Landsberg erecta (Ler), were used in parallel (Table  I). Representative photographs of plants used for gasexchange measurements (Supplemental Fig. S1) show that even mutants with high stomatal conductance (g st ) were healthy and nonwilted in our growth conditions.

ABA Signaling through PYR/RCAR Receptors Plays a Fundamental Role in Controlling Plant Steady-State g st
Whole-rosette g st varied severalfold among the studied mutants, ranging from 74 mmol m 22 s 21 in a triple lossof-function mutant of ABI1, HAB1, and PP2CA, abi1-2hab1-1pp2ca-1, to 683 mmol m 22 s 21 in the sextuple PYR/ RCAR loss-of-function mutant pyr1pyl1pyl2pyl4pyl5pyl8, abbreviated as 112458 ( Fig. 2A). The difference in g st between two wild types, Ler and Col-0, was 1.4fold. The g st values of 112458, the PYR/RCAR pentuple mutant pyr1pyl2pyl4pyl5pyl8 (12458), and dominant active PP2C mutants (abi1-1L and abi2-1L in Ler and abi1-1C and hab1 G246D in Col-0) were highest, followed by ABAdeficient aba1-1 and aba3-1 ( Fig. 2A). The hab1 G246D mutation is analogous to abi1-1 and abi2-1 mutations and causes resistance to ABA-dependent inhibition by PYR/ RCAR receptors, leading to strong ABA insensitivity (Robert et al., 2006;Dupeux et al., 2011). It should be noted that ABA biosynthesis mutants have reduced, but not abolished, ABA synthesis, with aba1-1 and aba3-1 still having 3% and 10% of the wild type ABA in stressed leaves, respectively (Rock and Zeevaart, 1991;Léon-Kloosterziel et al., 1996). Plants carrying loss-of-function mutations in SLAC1 and OST1 (slac1-3, ost1-2, and ost1-3) as well as in four PYR/RCAR receptors, pyr1pyl4pyl5pyl8 (1458) and pyr1pyl1pyl2pyl4 (note that we present data for this mutant in the Col-0 and Ler backgrounds, abbreviated as 1124C and 1124L, respectively), also exhibited significantly higher g st than the corresponding wild types. Gradual removal of PYR/RCAR receptor proteins had an increasing effect on g st . In two mutants with elevated ABA concentrations due to defective ABA catabolism (cyp707a1 and cyp707a3; Okamoto et al., 2009), only cyp707a1 had significantly reduced g st . OST1dependent phosphorylation of Ser-120 in SLAC1 is critical for channel activation in Xenopus laevis oocytes (Geiger et al., 2009) and in O 3 -induced plant stomatal response (Vahisalu et al., 2010). However, the g st of slac1-7, with an S120F point mutation, was similar to that of Col-0. CPK21 and CPK23 are other kinases shown to activate SLAC1 in X. laevis oocytes ; however, g st of cpk21 and cpk23 did not differ from that of wild-type plants ( Fig. 2A). Together, these results unequivocally demonstrate that mutations in the ABA signalosome have a major effect on the whole-plant stomatal conductivity and loss of water.
To address whether the differences in g st persisted after the application of 1 h of darkness, elevated CO 2 , and reduced humidity, we combined g st values of all mutants into one principal component analysis (PCA) axis. This PCA axis describes g st in stimuli-affected conditions. The ranking of genotypes by g st in stimuliaffected conditions (Fig. 2B) did not reveal any major differences from g st values in normal prestimuli conditions ( Fig. 2A): although there were some relocations (e.g. ost1-3 and 1124C), the groups of statistical significance remained unchanged. The result that plants with more open stomata remain more open even after receiving a signal to close may indicate either that stomatal closure is impaired due to the given mutation or that the absolute extent of stomatal closure induced by these stimuli is generally limited.

Data Analysis
Since the number of studied mutants was high, we determined two characteristics describing stomatal closure and enabling quantitative comparisons of different genotypes. First, we calculated changes of g st as (g st2 2 g st1 )/(t 2 2 t 1 ), as shown in Figure 3, A to C. Since t 2 2 t 1 is a similar time interval for all mutants (Fig. 3, A-C), this number describes the initial stomatal response to the applied factor (i.e. the magnitude of change in g st ). This characteristic can be effectively applied to all mutants. Stomatal closure in response to darkness, elevated CO 2 , and reduced air humidity of wild-type plants followed an exponential function (Fig. 3D). Thus, to provide a value describing the stomatal response kinetics of different genotypes, an exponential function was fitted to their stomatal closure responses and the maximum stomatal closure rate was calculated (Fig. 3D). However, several mutants either did not have clear stomatal responses to applied stimuli or their stomatal responses did not reach a stable phase within the time frame in which the experiments were carried out (e.g. slac1-3 and ost1-3 in Fig. 4, A and C, as examples, respectively). In such cases, fitting stomatal response to exponential function was not possible, and we interpreted this as an indication that the stomatal closure was affected due to the respective mutation. Data summarizing the results of exponential fitting are shown in Table II and Supplemental Figure S3.
For clarity and to indicate at which step CO 2 -, darkness-, reduced humidity-, and O 3 -induced stomatal closures diverge, we present the mutants of ABA signaling and SLAC1 activation from the bottom up, starting from SLAC1 anion channel regulation and moving stepwise to mutants defective in PP2C phosphatases, PYR/PYL proteins, and, finally, ABA biosynthesis and catabolism. Loss-of-function mutations in the SLAC1 anion channel and its main regulator, the protein kinase OST1, led to significantly impaired stomatal responses to darkness, high CO 2 , reduced air humidity, and O 3 , and the corresponding initial changes of g st were significantly lower Figure 2. A, There is a large variation in wholeplant steady-state stomatal conductance of plants with mutations in ABA signalosome. The average stomatal conductance values of three-to fourweek-old mutants and corresponding wild types. Significant differences (P , 0.05, n 5 6-46) are denoted with different small and capital letters for Col-0-and Ler-based mutants, respectively. B, Ranking of genotypes by their stomatal conductance after application of 1 h darkness, elevated CO 2 and reduced humidity derived from principal component analysis. Significant differences (P , 0.05, n 5 6-46) are denoted with different small and capital letters for Col-0 and Ler-based mutants, respectively. than in the wild type ( Fig. 4; Supplemental Fig. S2). In all genotypes, an initial sudden increase in g st was detected after the transition from humid to dry air. This is caused by a rapid increase in water evaporation from the epidermal cells and a concomitant decrease of their pressure on guard cells (Ivanoff, 1928). Due to this effect and the extremely slow closing response, the g st of OST1 loss-of-function mutants remained higher after 1 h in dry air compared with humid air (Fig. 4B;Supplemental Fig. S2B). Humidity-and darkness-induced stomatal closures were the only responses where significant differences between SLAC1 and OST1 loss-of-function plants were detected (Fig. 4, F and G).
The stomata of slac1-7 closed significantly less than those of Col-0 in response to all stimuli, confirming that the phosphorylation of Ser-120 was important for SLAC1 activation (Fig. 4;Geiger et al., 2009;Vahisalu et al., 2010). However, both the wild-type-like absolute g st ( Fig. 2A) and the clearly weaker phenotype in CO 2and O 3 -induced stomatal responses of slac1-7 compared with those observed in slac1-3 (Fig. 4, E and H) suggest that the phosphorylation of Ser-120 does not fully explain the activation mechanism of SLAC1. The maximum stomatal closure rate of slac1-7, derived by fitting the kinetics of CO 2 -induced stomatal closure with exponential function, yielded values lower than in the wild type (Supplemental Fig. S3B). These results confirm that, in addition to O 3 -induced stomatal closure (Vahisalu et al., 2010), SLAC1 Ser-120 is the target for OST1 in CO 2 -, reduced humidity-, and darknessinduced responses as well, but they also suggest that for full SLAC1 activation, phosphorylation of multiple Ser residues either by OST1 or in combination with other protein kinases shown to activate SLAC1 in X. laevis oocytes Brandt et al., 2012;Hua et al., 2012;Scherzer et al., 2012) is needed.
To address the role of CPK21 and CPK23, Ca 2+related protein kinases shown to activate SLAC1 in X. laevis oocytes , cpk21 and cpk23 plants were used. The stomata of cpk21 responded to the studied stimuli like the wild type, whereas the initial changes of g st of cpk23 were significantly reduced (Fig. 4, E-H), although much less than those of slac1-3 and ost1-3. Fitting the kinetics of stomatal closure with exponential function further confirmed that darkness-and humidity-induced stomatal closure is impaired in cpk23 (Table II; Supplemental Fig. S3A).
In conclusion, our results indicate that OST1-induced phosphorylation of SLAC1 is needed for rapid stomatal closure in response to all studied stimuli. Additionally, CPK23 is required, although to a minor extent.
PP2Cs Are Important for Rapid O 3 -and Humidity-Induced Stomatal Closure But Less Important for Darkness-and CO 2 -Induced Closure PP2Cs function as negative regulators in ABA-induced stomatal closure by inhibiting the OST1-and CPKinduced activation of SLAC1 (Geiger et al., 2009Brandt et al., 2012). In the presence of ABA, their activity is suppressed by PYR/RCAR receptors. We used plants carrying dominant active abi1-1, abi2-1, and hab1 G246D mutations that prevent ABA-dependent inhibition of PP2Cs by PYR/RCAR receptors. Humidity-and O 3induced stomatal responses and initial changes of g st were reduced in abi1-1C (Fig. 5, B, F, D, and H) and abi1-1L and abi2-1L (Supplemental Fig. S4, B, F, D, and H) as compared with their wild types, whereas in hab1 G246D , only the O 3 response was reduced (Fig. 5,D and H). Thus, for O 3 -and humidity-induced stomatal closure, the inhibition of PP2C activity is important.
The role of PP2Cs in CO 2 -and darkness-induced stomatal closure was less clear. The initial changes in g st of dominant abi1-1C and hab1 G246D mutants in the Col-0 background (Fig. 5, A, E, C, and G) and similarly those of abi1-1L and abi2-1L in the Ler background (Supplemental Fig. S4, A, E, C, and G) were generally wild type like. Similar results were obtained for plants with abi1-1 and abi2-1 mutations in the Col-0 and Ler backgrounds when analyzing elevated CO 2 -induced stomatal closure with a leaf gas-exchange analyzer (Supplemental Fig. S5). However, obtained patterns of stomatal closure could not be described with an exponential function in several cases, such as the darkness responses of abi1-1L and abi2-1L and the CO 2 responses of abi1-1C and abi2-1L (Table II), suggesting that the closure responses of these mutants were different from their wild types.
To further address the roles of ABI1 and ABI2 phosphatases for CO 2 -induced stomatal regulation, two Figure 3. A to C, Representative time courses of g st in response to darkness and CO 2 enrichment (A), decreased air humidity (B), and O 3 pulse (C) are given to illustrate the calculation of initial changes in stomatal conductance. D, An example of fitting the kinetics of stomatal closure with exponential function, where a describes the overall stomatal response and g is the value at which g st stabilized in response to stimuli. In case the stomatal closure significantly followed an exponential function, maximum stomatal closure rate was calculated as a 3 b. RH, Relative humidity.
additional experiments were performed. First, in a separate gas-exchange experiment, abi1-1L, abi2-1L, and Ler stomatal responses were tested within a wider range of CO 2 concentrations. Plants were first acclimatized under 50 mL L 21 CO 2 until stable g st values were reached (75-100 min). Thereafter, CO 2 concentration was increased stepwise to 100, 200, 400, and 800 mL L 21 in 30-min intervals. Such treatment induced clear stomatal closure in all three genotypes (Fig. 6A); however, when the stomatal closures caused by each additional step in CO 2 concentration were determined, differences between the genotypes emerged (Fig. 6B). There was no reduction in g st within the CO 2 range from 50 to 100 mL L 21 . Furthermore, from 50 to 200 mL L 21 , a decrease in g st was observed only in Ler. Change from 50 to 400 mL L 21 CO 2 caused a closure response in all genotypes; however, the decrease in g st was significantly smaller in abi2-1L than in abi1-1L, and this difference between abi1-1L and abi2-1L remained at 800 mL L 21 , where the CO 2 -induced decrease in g st was similar in Ler and abi1-1L but significantly lower in abi2-1L. Second, recent research showed that b-carbonic acid anhydrases function early in CO 2 -induced stomatal closure  and that bicarbonate (HCO 3 2 ) is an important intracellular signal that triggers the activation of S-type anion channels in Arabidopsis guard cells (Xue et al., 2011). To further address the roles of ABI1 and ABI2 in CO 2induced stomatal signaling, HCO 3 2 -induced activation of S-type anion currents was measured in abi1-1L and abi2-1L. Guard cell protoplasts from abi1-1L and abi2-1L displayed clearly reduced but still functional HCO 3 2 -induced activation of anion currents (Fig. 6, C-K).
In conclusion, PP2Cs are important for stomatal closure in response to reduced humidity and O 3 , and they also participate in darkness-and CO 2 -induced responses.
Removal of Six PYR/RCAR Receptor Proteins Impairs Plant Stomatal Responsiveness to O 3 , Reduced Humidity, Elevated CO 2 , and Darkness Gradual removal of PYR/RCAR receptor proteins increasingly impaired the whole-plant stomatal  Figure 3, A to C. Significant differences (P , 0.05; n = 5-50) are denoted with different lowercase letters. responsiveness to environmental factors; the sextuple 112458 PYR/RCAR mutant showed the strongest effect and displayed significantly impaired stomatal responses (Fig. 7, A-D) and reduced initial changes of g st due to all stimuli, except for elevated CO 2 (Fig. 7, E-H). Additionally, patterns of 112458 stomatal closure did not follow an exponential function in darkness, CO 2 , and humidity experiments (Table II), suggesting that the lack of these proteins modified the fast kinetics of stomatal closure in response to these stimuli. A separate gasexchange experiment with stepwise increases in CO 2 concentration also revealed that CO 2 -induced stomatal closure is reduced in 112458 compared with the wild type, particularly at lower CO 2 concentrations (Fig. 7, I and J). Thus, the lack of six PYR/PYL proteins significantly impaired stomatal closure due to all studied factors. Quadruple PYR/RCAR mutants displayed impaired stomatal responsiveness as well; however, here, the differences from the wild type depended on the applied stimuli and the combination of PYR/RCAR mutations. For example, O 3 -and humidity-induced initial changes of g st in quadruple 1124C (Fig. 7, F and H) and 1124L (Supplemental Fig. S6, F and H) mutants were significantly reduced, whereas 1458 quadruple and even 12458 pentuple mutants showed similar or even higher than wild-type closures in response to reduced humidity and O 3 (Fig. 7, F and H). These data indicate a functional diversity among PYR/RCAR proteins and suggest the importance of PYL1 for stomatal functioning. Recently, it was demonstrated that PYL1 played an important role in ABA-induced transcriptional response as well (Gonzalez-Guzman et al., 2012). Various combinations of triple loss-of-function PYR/PYL mutants, including pyr1pyl1pyl4, previously found to have higher steady-state g st values than the wild type (Gonzalez-Guzman et al., 2012), generally showed initial changes of g st that were wild type like or even larger than in the wild type (Supplemental Fig. S7). This suggests that a certain threshold of PYR/RCAR receptors is required in guard cells to trigger stomatal closure in response to environmental factors. Furthermore, the compensatory changes in the concentration/activity of other PYR/ RCARs in triple loss-of-function PYR/RCAR mutants can explain why their initial rates of stomatal closure were sometimes higher than in the wild type.

Elevated CO 2 , Reduced Air Humidity, Darkness, and O 3 -Induced Stomatal Closure of ABA Biosynthesis and Catabolism Mutants
Plants with mutations in ABA biosynthesis had weaker impairments of stomatal responses (Fig. 8) than those observed for plants impaired in ABA signaling (Figs. 4-7). The initial changes of g st in aba1-1 and aba3-1 did not differ from the Ler and Col-0 wild types, respectively (Fig. 8, E, G, and H). The only exception was significantly lower darkness-induced stomatal closure in aba1-1 than in its Ler wild type (Fig.  8G). Furthermore, the stomatal kinetics of aba1-1 did not follow an exponential function in darkness and CO 2 experiments. ABA is most likely essential for proper plant development; hence, no null mutants for ABA biosynthesis have been isolated. Accordingly, residual amounts of ABA in aba1-1 and aba3-1 (Rock and Zeevaart, 1991;Léon-Kloosterziel et al., 1996;Xie et al., 2006) could be sufficient to activate ABA signaling, since half-maximal inhibition of PP2Cs takes place at nanomolar ABA concentrations in the presence of PYR/RCARs (Szostkiewicz et al., 2010).
Surprisingly, the initial reductions in g st due to the reduced humidity of ABA biosynthesis mutants were significantly larger than those of the respective wild types (Fig. 8F). This result is confirmed by significantly larger maximum stomatal closure rates derived from exponential curve fitting of aba1-1 and aba3-1 (Supplemental Fig. S3C). This was an unexpected but not unique result: Assmann et al. (2000) found that aba1-1 plants showed a greater than wild-type stomatal response to an increase in leaf-air vapor pressure difference from 0.4 to 0.7 kPa. Since submission of the original version of this manuscript, it was reported that guard cell-autonomous ABA biosynthesis is important for low humidity-induced stomatal responses (Bauer et al., 2013). However, our study suggests the alternative model where basal low ABA in aba3-1 and Table II. Results of fitting exponential functions to stomatal closure patterns in response to darkness, elevated CO 2 , and reduced air humidity The first number shows how many closure responses could be described with an exponential function, whereas the second number shows the total number of experiments. Boldface values indicate cases where 50% of experiments did not follow an exponential function.

Genotype
Darkness Response Humidity Response

DISCUSSION
Rapid stomatal closure is one of the fastest responses in plant adaptation to sudden changes in environmental conditions. The finding that ABA perception by PYR/RCAR receptors leads to the inhibition of PP2C phosphatases has been a major breakthrough in plant science to understand ABA signaling and the regulation of stomatal aperture by ABA (Ma et al., 2009;Park et al., 2009). With these recent advances, the key question arises whether rapid changes in g st caused by physiological stimuli are affected by defined steps within the ABA signalosome in intact whole plants.
Here, we have addressed the relevance of the ABA signalosome ( Fig. 1) for the whole-plant steady-state g st and stomatal responses to darkness, reduced air humidity, elevated CO 2 , and O 3 (i.e. environmental factors). As this study used a large number of mutants and four different factors, we have summarized the descriptions of mutants into one table (Table I) and their stomatal responses to studied environmental factors into a simplified figure (Fig. 9B) to help readers follow the main results of the study.  signalosome: reduced ABA concentration, gradually reduced levels of functional PYR/RCAR proteins, the presence of dominant active PP2Cs, and the lack of functional OST1 and SLAC1 resulted in higher g st . Contrarily, reduced levels of functional PP2Cs in abi1-2hab1-1pp2ca-1 and higher ABA concentrations in cyp707a3 resulted in lower g st (Fig. 2). Previous gas-exchange experiments have also found that the mutants of the ABA signaling module have high g st (abi1-1 and abi1-2 [Leymarie et al., 1998a[Leymarie et al., , 1998b; aba1, abi1-1, and abi2-1 [Assmann et al., 2000]; and abi1-1, abi1-2, and ost1-3 [Xue et al., 2011]), whereas the reduced g st of triple lossof-function PP2C mutants (hab1-1abi1-2pp2ca-1 and hab1-1abi1-2abi2-2) and ABA catabolism mutants (cyp707a1 and cyp707a3) was detected previously by Rubio et al. (2009) and Okamoto et al. (2009), respectively. However, when the stomatal apertures from extracted epidermal fragments have been measured, a larger aperture of open stomatal mutants is not always evident (ost1-1 and ost1-2 [Mustilli et al., 2002]; hab1 G246D [Robert et al., 2006]; ost1-2 and abi1-1 [Siegel et al., 2009]; pyr1pyl1pyl2pyl4 ; and ost1-3 [Xue et al., 2011]). Earlier and more recent studies indicated that at least part of the stomatal responses to CO 2 and light depend on signals generated by the mesophyll (Lee and Bowling, 1992;Mott et al., 2008). Thus, it is expected that mesophyll or signals from the mesophyll may also play a role in determining the plant steady-state g st .
Nevertheless, it is obvious that plant steady-state g st is not determined only by the signal flow from ABA to OST1 and the SLAC1 anion channel. For example, the g st values of dominant active PP2C mutants were almost two times higher than those of OST1 and SLAC1 loss-of-function mutants. Guard cell plasma membrane H + -ATPases, activated by phosphorylation, provide the driving force for stomatal opening (for review, see Shimazaki et al., 2007). The basal and blue light-induced phosphorylation of the guard cell H + -ATPase was higher in abi1-1 and abi2-1 than in the wild type (Hayashi et al., 2011), indicating that the higher g st of dominant active PP2C mutants could be caused by their higher H + -ATPase activity. It remains to be established whether high H + -ATPase activity also explains the highest g st of plants lacking six PYR/RCAR receptors and the mechanistic connection between the ABA signalosome and the phosphorylation of guard cell H + -ATPases.   Figure 3, A to C. Significant differences (P , 0.05; n = 5-50) are denoted with different letters. I, Time courses of g st in response to stepwise changes in CO 2 from 50 to 800 mL L 21 in 112458 and Col-0 plants (n = 6). J, Changes in g st induced by each step of CO 2 .
Strongly reduced darkness-, reduced air humidity-, and O 3 -induced stomatal closure of 112458 together with its altered CO 2 response kinetics indicate that PYR/RCAR receptors influence the rapid initiation of stomatal closure by these stimuli. Furthermore, downstream components of the ABA signalosome (ABI2, OST1, and SLAC1) were required, since stomatal responses of these mutants were reduced (Fig. 9B). Only ABA biosynthesis mutants did not fit in, generally showing wild-type-like stomatal closure, but this can be explained by their residual ABA concentrations. Thus, rapid stomatal closure induced by reduced air humidity, darkness, elevated CO 2 , and O 3 involves PYR/ RCAR-dependent inhibition of PP2Cs, leading to the activation of OST1 and SLAC1.

Partial inhibition of HCO 3
2 -induced activation of S-type anion currents in abi1-1L and abi2-1L guard cells, together with the result that the obtained CO 2 -induced stomatal closure patterns of abi1-1C and abi2-1L could not be described with an exponential function, indicate that these dominant active phosphatases can affect the stomatal response to CO 2 ( Fig. 6; Table II). It is of particular interest that the delay in CO 2 -induced stomatal closure was clearly stronger in abi2-1L than in abi1-1L (Fig. 6B). Recently, a new regulator, GHR1, involved in ABA-and hydrogen peroxide-induced activation of SLAC1 was identified (Hua et al., 2012). GHR1, a receptor-like kinase preferentially localized in guard cell plasma membranes, was shown to activate SLAC1 anion currents in X. laevis oocytes (Hua et al., 2012). Interestingly, GHR1 is regulated by ABI2 but not by ABI1 (Hua et al., 2012). Thus, the differential responses of abi1-1L and abi2-1L could be a result of their different roles in the regulation of GHR1.
OST1 and SLAC1 Are Important for Stomatal Closure by All Four Stimuli, But There Are Additional Components Whose Exact Roles in Plant Stomatal Regulation Remain to Be Clarified CO 2 -and O 3 -induced stomatal closures were small, whereas darkness-and humidity-induced closures were delayed but still functional in SLAC1 loss-of-function plants (Negi et al., 2008;Vahisalu et al., 2008;Ache et al., 2010;Xue et al., 2011;this study). There are other anion channels that participate in stomatal closure together with SLAC1, including the membrane voltagedependent rapid-type anion channel QUAC1 (Meyer et al., 2010) and SLAH3, another slow-type anion channel that is activated in X. laevis oocytes via phosphorylation by CPK21 (Geiger et al., 2011). It remains to be determined why, then, SLAC1 has a vital role in regulating CO 2 -and O 3 -induced stomatal closure (Fig. 4, A and D), whereas in humidity and darkness responses, QUAC1 or possibly SLAH3 could partly replace SLAC1 function (Fig. 4, B and C). Perhaps since darkness and drought, with accompanying decreases in air humidity, are important environmental factors that have affected plants in the evolutionary time scale, plants have developed parallel signaling pathways and ion channels mediating rapid stomatal closure in response to these  (B). A, The emerging model suggests that in the case of O 3 , all components of the ABA signalosome are required to trigger stomatal closure. For reduced air humidity, we show that the ABA signalosome plays an important role; however, the presence of a parallel, ABA-independent pathway (marked as 1) was suggested by Assmann et al. (2000) and Xie et al. (2006). Darkness-induced stomatal closure is mediated by increased intercellular CO 2 concentration that activates anion channels and by inactivation of H + -ATPase (Roelfsema et al., 2002;Shimazaki et al., 2007). CO 2 -induced stomatal closure involves the activation of carbonic acid anhydrases that convert CO 2 to HCO 3 2 . The results presented in this study suggest that CO 2 -induced stomatal closure and HCO 3 2 -induced activation of S-type anion channels are partly controlled by the ABA signalosome. Question marks highlight that the nature of signal perception at the plasma membrane and signal transduction in the cytosol leading to the activation of the ABA signalosome remain to be addressed. B, Summary of stomatal responses to closure-inducing stimuli in the studied mutants. Mutants are presented in the order of phenotypic severity. H and D indicate air humidity and darkness, respectively. + indicates that either initial change in g st or curve fitting was different from the wild type, and ++ indicates that initial change in g st and curve fitting were both different from the wild type. The absence of a symbol indicates wildtype-like stomatal closure. Mutants that showed wild-type-like stomatal responses to all stimuli (148, cpk21, cyp707a1, and cyp707a3) are not listed here. [See online article for color version of this figure. ] factors. Furthermore, darkness-induced stomatal closure is accomplished through two different signaling pathways. First, it is mediated via phototropins and H + -ATPase (Shimazaki et al., 2007), since blue light, which activates phototropins, is part of visible light spectrum. Second, CO 2 is an intermediate signal in the darkness response; photosynthesis immediately stops in darkness, resulting in increased intercellular CO 2 concentration and activation of anion channels (Roelfsema et al., 2002). Thus, the partial darkness response of SLAC1 and OST1 loss-of-function mutants could be caused by signaling via phototropins and H + -ATPase, while the CO 2 -mediated signaling remains inactive.
It is also important to consider that in all studied ABA signalosome mutants, compensatory changes, either directly or indirectly related to the signalosome itself, could have taken place and affected the wholeplant stomatal response. For example, the regulation of PP2C activity can become less or more sensitive to ABA, depending on the PP2C:PYR/RCAR ratio (Szostkiewicz et al., 2010), representing compensation within the ABA signalosome. Besides guard cell anion channels, the ABA signalosome also regulates the activity of guard cell potassium channels (Armstrong et al., 1995;Sato et al., 2009). Recently, it was shown that plants with impaired SLAC1 have slowed stomatal opening in response to various stimuli and that this is caused by strongly reduced inward-rectifying K 1 channel activity of slac1 mutants (Laanemets et al., 2013). These unexpected phenotypes of slac1 turned out to be caused by higher cytosolic pH and Ca 2+ concentration  and increased Ca 2+ sensitivity of inward-rectifying K 1 channels in slac1 guard cells (Laanemets et al., 2013). These changes represent adaptive changes not directly related to the ABA signalosome and counteract the adverse effects of the slac1 mutation and allow the plant to maintain control over stomatal openness. Thus, while interpreting the results of this study, it is important to consider that some compensatory changes that also affect stomatal regulation have probably occurred in the studied mutants.
The calcium-dependent protein kinases CPK21 and CPK23 phosphorylate and activate SLAC1 similarly to OST1 . The stomata of CPK23 lossof-function plants showed slightly reduced responses to all environmental factors. Interestingly, Ma and Wu (2007) found that CPK23 acts as a positive regulator for stomatal opening: the stomatal apertures of epidermal peels were significantly decreased in the cpk23 loss-offunction mutant, resulting in enhanced drought and salt tolerance. In this study, whole-plant g st in the cpk23 mutant was similar to that in the wild type ( Fig.  2A). Phosphorylation by CPK21 was found to activate both SLAC1  and SLAH3 (Geiger et al., 2011). However, no changes in stomatal responses to the studied stimuli were found in the cpk21 mutant. This may not be not surprising, considering that the family of Ca 2+ -dependent protein kinases is large, with possible redundancy in their function (Cheng et al., 2002). As an example, CPK3 and CPK6 were found to participate in ABA-and Ca 2+ -dependent regulation of guard cell S-type anion channels and stomatal closure (Mori et al., 2006). Recently, CPK6 was shown to strongly activate SLAC1-mediated anion currents in X. laevis oocytes and to allow the functional reconstitution of ABA activation of SLAC1 (Brandt et al., 2012). Interestingly, Ser-59 in the SLAC1 N terminus was phosphorylated by CPK6, and this phosphorylation is essential for SLAC1 activation (Brandt et al., 2012). Ser-59 was earlier shown to be phosphorylated also by OST1 (Vahisalu et al., 2010). The strong stomatal phenotypes of OST1 loss-of-function plants in the presence of many alternative kinases (CPK3, CPK6, CPK21, and CPK23) activating SLAC1 suggests that interaction of the Ca 2+ -dependent and Ca 2+ -independent pathways requires further investigation.
In conclusion, the signaling pathways of different stomatal closure-inducing factors converge at OST1 and SLAC1. In darkness-, O 3 -, and reduced air humidityinduced stomatal closure, signaling through PYR/ RCAR receptors plays an important role. In response to elevated CO 2 , the ABA signalosome was partially involved, and the presence of a parallel, yet to be identified signaling pathway that activates OST1 is possible.

Plant Material and Growth Conditions
Arabidopsis (Arabidopsis thaliana) seeds were planted in soil containing 4:3 (v/v) peat:vermiculite and grown through a hole in a glass plate covering the pot as described (Kollist et al., 2007). Soil moisture was kept at 60% to 80% of maximum water capacity. Plants were grown in growth chambers (AR-66LX and AR-22L; Percival Scientific) in a 12/12-h photoperiod, 23°C/ 18°C temperature, 150 mmol m 22 s 21 light, and 70% to 75% relative humidity. For gas-exchange experiments, we used plants with a total rosette area between 5 and 15 cm 2 . This corresponds to 21-to 25-d-old plants for most mutant lines. However, some mutants (aba1-1, aba3-1, abi1-1C, and hab1 G246D ) had slower growth rates; thus, older plants (26-32 d old) were analyzed. The full list of the mutants used is given in Table I. Mutants were obtained from the European Arabidopsis Stock Centre (www.arabidopsis.info) and from Sean Cutler (114, 1124C, and 1124L). The mutants cpk21 (GABI_322A03), cpk23 (SALK_007958; Geiger et al., 2010), cyp707a1 (SALK_069127; Okamoto et al., 2009), and cyp707a3 (SALK_078173; Okamoto et al., 2009) were confirmed to be homozygous using PCR. The cpk21 knockout mutant was verified to lack full-length transcripts for CPK21 using reverse transcription-PCR (Supplemental Fig. S8).

Whole-Rosette g st Measurements
The Arabidopsis whole-rosette rapid-response gas-exchange measurement device was described previously (Kollist et al., 2007;Vahisalu et al., 2008). Plants were inserted into the device, and the treatments started about 2 h later, when g st had stabilized. Photographs of plants were taken before the experiment, and rosette leaf area was calculated using ImageJ 1.37v (National Institutes of Health). g st for water vapor was calculated with a custom-written program as described by Kollist et al., (2007).
In light-dark transition experiments, the g st values were first measured in light, and then darkness was applied for 60 min by covering the measuring cuvettes. In CO 2 enrichment experiments, plants were kept in ambient CO 2 concentration (400 mL L 21 ) until g st was stable, and then CO 2 concentration was increased to 800 mL L 21 for 60 min. In darkness and elevated CO 2 experiments, we also followed the reopening of stomata when light and ambient CO 2 were restored. In O 3 experiments, plants were exposed to 350 to 450 nL L 21 O 3 for 3 min and kept in measuring cuvettes for 60 min after exposure. In Plant Physiol. Vol. 162, 2013 1665 Environmental Factor-Induced Stomatal Closure reduced humidity experiments, plants were kept in humid air (relative humidity = 60%-80%), air humidity was abruptly reduced about two times (relative humidity = 30%-40%), and g st was followed for the next 56 min. We present the time-resolved stomatal responses to stimuli in absolute units. Although the course of stomatal reopening in light and ambient CO 2 is not discussed, it is presented in the figures for darkness and CO 2 experiments. In order to provide a quantitative value for the initial stomatal responsiveness to applied stimuli, we calculated changes of g st as described in Figure 3, A-C. Furthermore, we fitted the patterns of g st values in response to darkness, elevated CO 2 , and reduced air humidity with exponential functions [g st = a 3 exp (b 3 time) + g]. Fit was accepted as significant when both a and b were significantly different from 0 at P , 0.1; in this case, we calculated maximum stomatal closure rate as a 3 b (Fig. 3D).

Electrophysiology
Arabidopsis guard cell protoplasts were isolated as according to Siegel et al. (2009). Whole-cell patch-clamp recordings were performed as described previously (Pei et al., 1997). For S-type anion current recordings (Schroeder and Keller, 1992), the pipette solution contained 150 mM CsCl, 2 mM MgCl 2 , 6.7 mM EGTA, 5.86 mM CaCl 2 (2 mM intracellular [Ca 2+ ]), 5 mM Mg-ATP, and 1 mM HEPES-Tris, pH 7.1. The bath solution contained 30 mM CsCl, 2 mM MgCl 2 , 1 mM CaCl 2 , and 10 mM MES-Tris, pH 5.6. Osmolalities of all solutions were adjusted to 485 mmol kg 21 for bath solution and 500 mmol kg 21 for pipette solution by the addition of D-sorbitol. The membrane voltage was stepped from +35 mV to 2145 mV with 230-mV decrements, and the holding potential was +30 mV. Liquid junction potential was determined using Clampex 10.0. No leak subtraction was applied for all current-voltage curves. Steadystate currents were the average currents during the last 500 ms of voltage pulses. For HCO 3 2 activation of S-type anion currents, 13.5 mM total HCO 3 2 (equivalent to 11.5 free intracellular [HCO 3 2 ] and 2 mM free intracellular [CO 2 ]) was added freshly in the pipette solution. The details were described previously (Xue et al., 2011).

Statistical Analysis
Statistical analyses were performed with Statistica, version 7.0 (StatSoft). ANOVA (General Linear Models) was used to assess the effect of genotype on g st , initial changes in g st , and maximum stomatal closure rates, and comparisons between individual means were done using Fisher's LSD test. Data were ln transformed when necessary. All effects were considered significant at P , 0.05. Exponential fitting of stomatal closure responses due to darkness, elevated CO 2 , and reduced air humidity was done with nonlinear least-squares model estimation of Statistica (Gauss-Newton estimation method). PCA was used to combine the values of g st after 1 h in darkness, elevated CO 2 , and reduced humidity into one PCA axis describing whole-plant g st after application of given stimuli.

Supplemental Data
The following materials are available in the online version of this article.
Supplemental Figure S1. Representative photographs of mutants and wild types.
Supplemental Figure S2. Time-courses of g st in response to stimuli and initial changes in g st values of ost1-2 and Ler.
Supplemental Figure S3. Maximum stomatal closure rates derived from exponential fitting of stomatal closure responses.
Supplemental Figure S4. Time-courses of g st in response to stimuli and initial changes in g st values of abi1-1L, abi2-1L and Ler.
Supplemental Figure S5. Time-courses of leaf g st in response to elevated CO 2 in abi1-1 and abi2-1 in Col-0 and Ler backgrounds.
Supplemental Figure S6. Time-courses of g st in response to stimuli and initial changes in g st values of 1124L and Ler.
Supplemental Figure S7. Time-courses of g st in response to stimuli and initial changes in g st values of triple loss-of-function mutants of PYR/ RCARs.
Supplemental Figure S8. The cpk21 T-DNA mutant is a transcriptional knockout.