ABA signalling and metabolism are not essential for dark-induced stomatal closure but affect response speed

Stomata are microscopic pores that open and close, acting to balance CO2 uptake with water loss. Stomata close in response to various signals including the drought hormone abscisic acid (ABA), microbe-associated-molecular-patterns, high CO2 levels, and darkness. The signalling pathways underlying ABA-induced stomatal closure are well known, however, the mechanism for dark-induced stomatal closure is less clear. ABA signalling has been suggested to play a role in dark-induced stomatal closure, but it is unclear how this occurs. Here we investigate the role of ABA in promoting dark-induced stomatal closure. Tracking stomatal movements on the surface of leaf discs we find, although steady state stomatal apertures are affected by mutations in ABA signalling and metabolism genes, all mutants investigated close in response to darkness. However, we observed a delayed response to darkness for certain ABA signalling and metabolism mutants. Investigating this further in the quadruple ABA receptor mutant (pyr1pyl1pyl2pyl4), compared with wild-type, we found reduced stomatal conductance kinetics. Although our results suggest a non-essential role for ABA in dark-induced stomatal closure, we show that ABA modulates the speed of the dark-induced closure response. These results highlight the role of ABA signalling and metabolic pathways as potential targets for enhancing stomatal movement kinetics.

www.nature.com/scientificreports/ darkness 8 . A MEK1/MPK6 signalling cascade activated by H 2 O 2 (produced by RBOHD and RBOHF) and culminating in the production of NO (by NIA1) has also been shown to be required for dark-induced closure 9 . Studies have linked ABA signalling to dark-induced closure, but the precise way in which it is involved in regulating dark-induced closure is unclear. Microarray data have shown components of the ABA signalling pathway undergo transcriptional regulation in response to darkness, however these changes are likely to reflect longer term adaptation rather than the short term closure 10 . Additionally, another study has shown that a selection of ABA receptor mutants (pyr1pyl1pyl2pyl4, pyr1pyl4pyl5pyl8, pyr1pyl2pyl4pyl5pyl8, pyr1pyl1pyl2pyl4pyl5pyl8) all show increased stomatal conductance under light and dark conditions. However, when comparing the change in stomatal conductance from light to dark conditions all of the previously mentioned ABA receptor mutants (except the strongest mutant, pyr1pyl1pyl2pyl4pyl5pyl8) show changes in stomatal conductance similar to wild type. Similarly, mutations within PP2C phosphatases (downstream negative regulators of ABA signalling) and ABA degradation mutants, affect stomatal conductance without preventing responses to darkness. In the ABA biosynthesis mutants, aba1-1 and aba3-1, stomatal conductance is also increased, however both mutants still respond to a dark (although this appears weakened in aba3-1) 11 . In addition, stomatal aperture changes in the PP2C mutants abi1 and abi2 show reduced responses to darkness 12 . This suggests a situation where ABA signalling may make a contribution to dark-induced stomatal closure, however, it also suggests that ABA has more general effects on stomatal apertures regardless of light or dark conditions.
Here we investigate how defects in ABA signalling and metabolism affect stomatal response to darkness and light. We analyse the movement of stomata through direct measurements on leaf discs and through monitoring changes in stomatal conductance. We find evidence that ABA signalling is not essential for dark-induced closure. However, we provide additional evidence that defects in ABA signalling and metabolism affect the timing of stomatal responses to both light and darkness. Overall, we conclude that ABA signalling does not play a major role in mediating dark-induced closure but does play a role in modulating the speed of closure.

ABA signalling and biosynthesis mutant responses to darkness.
To explore whether mutations in ABA biosynthesis, degradation or signalling genes affect stomatal responses to darkness, stomatal movements were measured in leaf discs from ABA metabolism and signalling mutants. Aperture measurements taken over a 2 h dark treatment time course, with independent leaf disc measurements at each timepoint, allowed for detection of trends in stomatal responses. Of the 14 member ABA receptor family, quadruple and sextuple ABA receptor mutants (pyr1pyl1pyl2pyl4 13 -q1124 and pyr1pyl1pyl2pyl4pyl5pyl8 14 -s112458) were used. The ABA biosynthesis double mutant nced3nced5 15 (nced3/5-a double mutant in the NCED3 and 5 genes which catalyse the first committal step in ABA biosynthesis, thought to be the rate limiting step under drought conditions [15][16][17] and mutants within 2 genes involved in rapid ABA activation from inactive glucose esters (bg1 and bg2 18 ) were used. Additionally mutants in ABA hydroxylation genes (cyp707a1 and cyp707a3 19 ) involved in ABA catabolism were used (exact mutant accession codes are shown in the methods section).
Delays in stomatal closure following dark treatment were observed for the ABA receptor mutants q1124 and s112458, the ABA biosynthesis mutant nced3/5, and the ABA activation mutant bg1, which all show no significant change in stomatal aperture, compared with wild type, after 30 mins dark treatment (Fig. 1a-c). No delays in dark-induced stomatal closure were observed in the ABA activation mutant bg2 or the ABA catabolism mutants cyp707a1 or cyp707a3 (Fig. 1c,d). Absolute changes in stomatal aperture for each mutant are shown in Fig. S1. The absence of a delay phenotype in the bg2 mutant could be due to the difference in subcellular location and/ or reduced activity of the BG2 protein compared with BG1 18 . Additionally, mutations in CYP707A1 and 3 genes lead to increased levels of ABA 19,20 and increased ABA signalling activity (the opposite of what occurs in ABA biosynthesis and signalling mutants), explaining the lower starting apertures (as observed previously 19 ) and potentially the absence of delay.
Stomatal conductance responses to darkness. The stomatal conductance responses to darkness applied at midday were also measured in the following genotypes; Col-0, q1124, bg1, cyp707a1, and cyp707a3. Due to the greatly reduced leaf size phenotype of the s112458 and nced3/5 mutants, stomatal conductance could not be recorded. Focusing on dark-induced decreases in stomatal conductance, it is evident this response occurs much faster than dark-induced closure in leaf discs (Fig. 1). The absolute and relative stomatal conductance values are shown for the mutants in Fig. 2a, c, e and b, d, f respectively. Here, delayed responses are only noticeable for the q1124 receptor mutant (Fig. 2a,b). All mutants appear to decrease their stomatal conductance by around 50% in response to darkness. The difference in the periods of delays observed in leaf discs and in measurements of stomatal conductance may reflect that stomatal conductance responses are faster than changes in stomatal aperture measured on leaf discs, similar to observed differences in stomatal movement when comparing responses to red light in epidermal peels and intact leaves 21 . Stomatal conductance responses generally reach their maximum by 25 mins (Figs. 2, 4, 5, 6) whereas leaf disc stomatal aperture responses reach their maximum by 120 mins (Figs. 1, 7). The slower movements on leaf discs may allow for more subtle mutant phenotypes to www.nature.com/scientificreports/ be identified. Additionally, when tracking the intercellular leaf CO 2 concentration (C i ) over the course of the experiments a rapid increase is seen upon dark treatment, that then rapidly decreases upon reintroduction to www.nature.com/scientificreports/ light, before returning to levels comparable to those before the initial dark treatment (Fig. 3). This build-up of C i may contribute to the stomatal conductance response, however this requires further investigation.
Further analysis of q1124 stomatal conductance responses to darkness. For Col-0 and q1124, the stomatal conductance responses were measured when whole plants were placed in darkness 4-5 h after dawn. Leaves were clamped into the gas analyser leaf cuvette, 2 h prior to the onset of darkness. After 1 h of darkness plants were reintroduced to light for a further hour. q1124 shows higher stomatal conductance (at time 0-p = 0.00231) and stomatal conductance is reduced to a lesser extent than wild type in darkness. This is true in absolute (Fig. 4a) and relative (Fig. 4b) terms. The time taken for q1124 to reach half of its total stomatal conductance response to darkness (darkness half response time- Fig. 4c) and to light (light half response time- Stomatal conductance in Col-0 and q1124 was also measured at dusk. Unlike the response to darkness measured at midday, where plants were plunged into darkness, the onset of dusk was marked with a 15 min transition from light to dark. The absolute stomatal conductance of Col-0 and q1124 is shown in Fig. 5a. It is clear that both genotypes respond to dusk. Because of the differences in absolute initial stomatal conductance, relative stomatal conductance were calculated. The data in Fig. 5b suggest that the speed of stomatal conductance change in the q1124 ABA quadruple receptor mutant is reduced compared with Col-0. The difference between Col-0 and q1124 is less pronounced than that observed in Fig. 4, yet measurement of dusk half response times supports a slower stomatal conductance response of q1124 to darkness (p = 0.00607).
The effect of mutations in ABA biosynthesis, signalling and activation on stomatal responses to light. In addition to the delay in dark-induced stomatal closure and slower stomatal conductance responses observed in q1124, the data in Figs. 2 and 4 suggest that there might also be defects in the light-induced opening response of the quadruple receptor mutant. Stomatal conductance of q1124 was measured over the dawn period. Here, similarly to dusk, the onset of dawn was marked with a 15 min transition period from dark to light. Absolute and relative stomatal conductance values are presented in Fig. 6a, b respectively. Similar to the response observed at dusk, in comparison to wild type, the q1124 mutant shows increased absolute stomatal conduct- The stomatal movements of q1124, nced3/5 and bg1 mutants were also analysed in leaf discs. Here, leaf discs were harvested pre-dawn under green light, incubated in the dark for 2 h, before being transferred to light. The apertures were monitored over a 2 h time course. Figure 7 shows both the absolute and relative change in stomatal aperture for the three mutants. It is evident that both q1124 and nced3/5 have significantly increased stomatal apertures at 0 mins (p < 0.0001, p < 0.0001 respectively), whereas bg1 is similar to Col-0. This makes comparisons between the absolute stomatal apertures of Col-0, q1124 and nced3/5 more difficult to interpret. However, when analysing the absolute change in stomatal aperture for both q1124 and nced3/5 there are no initial significant differences at 30 mins, but by 120 mins there are significantly reduced responses (p < 0.005, p < 0.0001 respectively).
Unlike q1124 and nced3/5, the bg1 mutant shows a response analogous to that observed when plants are placed in darkness (except instead of an initial delay in dark-induced stomatal closure, here a delay in light-induced stomatal opening is observed). bg1 shows a significantly weakened response at 30 mins (p < 0.0005), before the bg1 mutant eventually catches up to Col-0 by 60 and 120 mins.

Discussion
ABA is well known as a regulator of seed dormancy and plant responses to drought including reductions in stomatal aperture and inhibition of light-induced opening 22 . Additionally, evidence is emerging supporting roles for basal ABA signalling under non-stress situations 23 . On a molecular level, a subset of ABA receptor family members (subfamily 1) are known to activate downstream signalling under basal amounts of ABA 24 and mutation of ABA signalling and metabolism components alters plant growth and development under non stress conditions 25,26 . Here we explore the role of ABA and ABA signalling in stomatal responses to the onset of light and darkness.
The mechanisms behind light-induced stomatal opening are relatively well understood. Stomatal opening in response to light is driven by the activity of plasma membrane H + ATPases. This generates a proton gradient across the guard cell plasma membranes, resulting in membrane hyperpolarization leading to the influx of cations and  www.nature.com/scientificreports/ anions, changes in guard cell turgor pressure, and ultimately the opening of stomata. Blue and red light promote stomatal opening via independent pathways. Blue light-induced opening is predominantly initiated via activation of phototropin photoreceptors within the guard cell 27 , whereas red light-induced opening is dependent on photosynthetic electron transport 28,29 . When Arabidopsis plants are moved from light to dark their stomata close, however, the mechanisms behind dark-induced stomatal closure are less clear 30 . Studies have linked ABA signalling to dark-induced closure, but it is unclear to what extent ABA signalling is required for this process 4,10-12 .
Here, we present evidence that supports a non-central role for ABA metabolism and ABA signalling in darkinduced stomatal closure. In leaf discs we observe that all ABA signalling and metabolism mutants analysed were able to respond to darkness (Fig. 1). Compared with wild type, ABA receptor mutants (q1124 and s112458) and the ABA biosynthesis mutant (nced3/5) showed increased stomatal apertures before treatment whereas ABA degradation mutants (cyp707a1 and cyp707a3) showed decreased stomatal apertures. This is in line with previous reports, namely that, defects in ABA signalling and production lead to increases in steady state stomatal apertures and transpiration, whereas defects in ABA degradation lead to the opposite 11,14,15,20 . This shows links between ABA, ABA signalling and the regulation of stomatal apertures under non stress conditions. However, following dark treatment cyp707a1 and cyp707a3 mutants close to the same extent as wild type. On the other hand, the stomata of q1124, s112458, and nced3/5 mutants all remain more open than wild type. These results are similar to those observed for stomatal conductance in ABA receptor mutants (q1124/pyr1pyl1pyl2pyl4, pyr1pyl4pyl5pyl8, pyr1pyl2pyl4pyl5pyl8) in Merilo et al. 2013 with one exception. Whereas Merilo et al. 2013 report no response to darkness, here we report that the stomata of s112458 mutant are still able to respond to darkness, but to a lesser extent than wild type. The strong ABA biosynthesis mutant nced3/5 also behaves similarly to the aba1-1 and aba3-1 biosynthesis mutants in Merilo et al. 2013. nced3/5 shows significantly more open stomata throughout the experiment and responds to darkness, although to a lesser extent than wild type. Additionally, we report ABA signalling (q1124 and s112458) and biosynthesis mutants (nced3/5 and bg1) show a delay in closure, with no significant change in stomatal aperture following 30 min of dark treatment, when measured in leaf discs. In www.nature.com/scientificreports/ mutants (aba1-1 and aba3-1), where stomatal conductances are increased but still show responses to darkness. This suggests a non-central role for ABA and ABA signalling in dark-induced closure.
Exploring this further we find that the time taken for q1124 mutant to reach its maximum stomatal conductance response is increased compared to Col-0 (Figs. 2,4,5,6). This is observed when darkness and light are applied during the middle of the day and also at the dawn and dusk transition periods. This suggests that although ABA signalling is not essential for dark-induced closure, it appears to be involved in increasing the speed of stomatal responses to darkness and light. It should be noted there is a difference between the time taken for stomatal apertures to show maximum responses (Fig. 1) and stomatal conductance responses to reach their maximum (Figs. 2, 4, 5, 6), with stomatal movement on leaf discs appearing slower than that of changes in conductance. The reasons for this are unclear but likely due to differences between the systems of a leaf disc and an attached leaf.
Manipulating the speed of stomatal responses has been shown to increase biomass accumulation 31 and may improve key plant processes such as photosynthetic carbon assimilation and water use efficiency 32 , suggesting it could be a key target for plant breeders. How defects within ABA metabolism and signalling are affecting the speed of stomatal responses is currently unclear but may stem from altered amounts or activities of further signalling components and/or ion channels within plant cells. Additionally, exogenous ABA application has been reported to alter stomatal kinetic responses to changes in light conditions in gymnosperms, suggesting roles for ABA in modulating stomatal kinetics across taxa 33 .
At the onset of darkness there is a rapid increase in CO 2 within the leaf (C i ), which rapidly decreases upon the re-introduction of light (Fig. 3). As photosynthetic CO 2 fixation ceases in darkness an increase in C i is not surprising. However, the role of C i in driving stomatal responses is unclear. Some studies suggest that under changing light conditions C i allows interaction between photosynthetic assimilation rate (A) and stomatal conductance, making C i a potential candidate for coordinating mesophyll and stomatal responses to light 34 . However, other studies show that stomatal conductance responses to changes in light still occur when C i is kept constant and in mutants where C i is increased 28,35 . This has led to the suggestion that other signals, not C i , are involved in coordinating stomatal responses with photosynthetic activity 1 . Regardless, the increase of C i in response to darkness is highlighted here (Fig. 3) but its contribution to stomatal responses to changes in light is beyond the scope of this study. www.nature.com/scientificreports/ Similar to our conclusions regarding the role of ABA in guard cell dark signalling, it has been reported that CO 2 -induced stomatal closure proceeds through an ABA-independent pathway downstream of OST1 and that basal ABA signalling enhances CO 2 -induced closure 36,37 . However, there is disagreement as to the precise role of ABA in stomatal responses to CO 2 , as other studies suggest that ABA and ABA signalling are required for elevated CO 2 induced stomatal closure 38 . Most recently, a study has shown ABA catabolism plays a role in regulating stomatal responses to changes in CO 2 concentration both on a physiological and developmental scale 39 .  www.nature.com/scientificreports/ In conclusion, our results using ABA signalling and metabolism mutants show that stomatal conductance and stomatal apertures decrease in response to darkness. The differences between the mutants and the wild type were reflected in the slower rates of closure and stomatal conductance changes found in the mutants. While our results do not support a primary role for ABA in the events underlying dark-induced stomatal closure, we find a role for this hormone in modulating the speed of reaction. This highlights the role of ABA in regulating stomatal aperture/transpiration under non-stress conditions.

Methods
Plant material and growth conditions. Arabidopsis thaliana was grown in a 3:1 all purpose compost (Sinclair): silver sand (Melcourt) mixture. Seeds were stratified in the dark for 48 h at 4 °C. Plants were grown under 120 µmol m −2 s −1 white light in short day conditions with a 10 h photoperiod, 22/20 °C day/night temperatures, and 70% relative humidity in a Snjider Labs Micro Clima-Series High Specs Plant Growth Chamber.
Stomatal aperture bioassays. Experiments were performed on 5 week old plants. Stomatal apertures were measured using leaf discs (4 mm in diameter). For dark-induced closure leaf discs were harvested 2 h after dawn, and incubated in petri dishes containing 10/50 buffer (10 mM MES/KOH, 50 mM KCl, pH 6.2) at 22 °C and illuminated with 120 μmol m −2 s −1 white light (Crompton Lamps 13 W white) for a further 2 h. Leaf discs were transferred to 10/50 buffer at 22 °C in darkness. Stomatal apertures were measured over a 2 h time course, at 0, 30, 60 and 120 min of dark treatment. A set of control leaf discs were kept in the light for 120 min over the same period (120 min L). For light-induced opening leaf discs were harvested prior to dawn under green light. Leaf discs were incubated in 10/50 buffer at 22 °C in darkness for 2 h before transfer to 120 mmol m −2 s −1 light. Stomatal apertures were measured over a 2 h time course at 0, 30, 60 and 120 min of light treatment. A set of control leaf discs were kept in the dark for 120 min over the same period (120 min D). For measurement of stomatal aperture leaf discs were imaged using a Leica DMI6000 B inverted microscope and apertures measured using ImageJ (FIJI). For each experimental repeat 30 apertures (over three individual plants) were measured for each treatment, in total over all repeats this amounts to 90 apertures per treatment per genotype. Data was statistically analysed using 2-way ANOVA with Tukey multiple comparison tests.

Stomatal conductance measurements.
Experiments were performed on 6-8 week old plants. A GFS3000 IR gas analyser (Walz) fitted with a 2.5 cm 2 leaf cuvette was used to measure transpiration. The leaf cuvette was set to 16,000 ppm H 2 O, 400 ppm CO 2 , 22 °C. Flow was set to 750 μmol s −1 and impeller speed was set to 7. For darkness applied at midday mature leaves were placed in the cuvette and left to equilibrate for 2 h in a plant growth cabinet under 120 μmol m −2 s −1 white light. Darkness was imposed for 60/120 min (Figs. 4 and 2 respectively), before reintroduction to 120 μmol m −2 s −1 white light. For dusk/dawn measurements mature leaves were placed in the cuvette at least 2 h prior to the onset of dusk. Plants were left in the cuvette throughout the night until 2 h post dawn the following day. Stomatal conductance half response times were determined by identifying the time required for transpiration to reach half of the total response over the experiment. Data was analysed using 1-way ANOVAs. R (version: 4.0.2, url: https ://www.r-proje ct.org/) was used to perform statistics and the package ggplot2 used to generate figures 42 .