Arabidopsis guard cell CO2/HCO3− response mutant screening by an aequorin-based calcium imaging system

Background The increase in atmospheric CO2 is causing a number of changes in plant growth such as increases in leaf area and number, branching, plant size and biomass, and growth rate. Despite the importance of stomatal responses to CO2, little is known about the genetic and molecular mechanisms that mediate stomatal development and movement in response to CO2 levels. Deciphering the mechanisms that sense changes in CO2 and/or HCO3− concentration is critical for unraveling the role of CO2 in stomatal development movement. In Arabidopsis, CO2-induced stomatal closure is strongly Ca2+-dependent. To further dissect this signaling pathway and identify new components in the CO2 response pathway, we recorded [Ca2+]cyt changes in mutagenized Arabidopsis leaves and screened for mutants with abnormal guard cell behavior in response to CO2/HCO3−. Results We observed that 1 mM HCO3− induces [Ca2+]cys transient changes in guard cells and stomatal closure both in light and darkness. The changes in [Ca2+]cys induced by HCO3− could be detected by an aequorin-based calcium imaging system. Using this system, we identified a number of Arabidopsis mutants defective in both [Ca2+]cyt changes and the stomatal response to CO2/HCO3−. Conclusions We provide a sensitive method for isolating stomatal CO2/HCO3− response genes that function early in stomatal closure and that have a role in regulating [Ca2+]cyt. This method will be helpful in elucidating the Ca2+-dependent regulation of guard cell behavior in response to CO2/HCO3−.


Background
The stomata, which are formed by pairs of guard cells, can be considered the gas-exchange valves of plants. Stomatal aperture is regulated by several factors including phytohormone levels, carbon dioxide (CO 2 ) concentration, humidity, light, and pathogens.
A higher ambient CO 2 concentration increases leaf intercellular CO 2 concentration and mediates stomatal closure in plants, whereas a lower CO 2 concentration triggers stomatal opening. CO 2 influences not only the stomatal response, but also the number of stomata per unit leaf. This number is decreasing due to the long-term effect of continuing CO 2 concentration increases [1].
Despite the importance of stomatal responses to CO 2 , little is known about the genetic and molecular mechanisms mediating stomatal development and movement in response to elevation in CO 2 . CO 2 levels have been increasing steadily, and it is estimated that atmospheric CO 2 will reach 550 ppm in 2050 compared with 400 ppm presently [2], so it is increasingly urgent to discover

Open Access
Plant Methods *Correspondence: maxiaonan@henu.edu.cn; bailing@henu.edu.cn 1 State Key Laboratory of Crop Stress Adaptation and Improvement, School of Life Sciences, Henan University, Kaifeng 475004, China Full list of author information is available at the end of the article the underlying mechanisms of guard cell regulation in response to CO 2 levels. CO 2 sensing in animals is mainly linked to α-carbonic anhydrases (α-CAs) [3], which are also important for CO 2 perception in fungi [3,4]. Carbonic anhydrases (CAs) can accelerate the conversion of CO 2 into HCO 3 − and H + , which in turn induce related responses. In plants, CO 2 also can be converted into HCO 3 − and H + by anhydrases [5]. The key question in understanding stomatal movement in response to CO 2 is the mechanism for perception of changes in CO 2 and/or HCO 3 − concentration. Despite the importance of anhydrase enzymes in CO 2 perception in mammalian and fungal systems [3,4], no orthologous α-CAs has been identified in plants.
There are six β-CAs in Arabidopsis thaliana. CA1-and CA4-related stomatal movements were controlled by CO 2 in guard cells, whereas a ca1 ca4 double mutant exhibited insensitive stomatal movement response to CO 2 [6]. Expression of a mammalian α-CA in the ca1 ca4 double mutant restored the stomatal response to CO 2 , implying that CA-mediated CO 2 catalysis to HCO 3 − and H + in guard cells is the key step for transmission of the CO 2 signal [6].
Through isolation and analysis of genetic mutants, a number of proteins have been identified that function in CO 2 -controlled stomatal movement, including the SLAC1 anion channel [7,8], the PATROL1 Munc 13-like protein [9], the AtALMT12/QUAC1 R-type anion channel [10], and the RHC1 MATE transporter [11]. The characterization of these proteins has contributed to our understanding of the mechanisms of CO 2 -regulated guard cell behavior. For example, transporter protein RHC1 acts as a bicarbonate sensor, and the high-CO 2induced stomatal closure mediated by RHC1 is controlled by inhibition of HT1 (HIGH LEAF TEMPERATURE1) activity [11,12].
HT1 is regarded as a negative regulator in the CO 2 signaling pathway: it functions by promoting phosphorylation of OST1 and thus inhibiting its kinase activity [11]. Furthermore, OST1 protein kinase has been proved essential for high-CO 2 -induced stomatal closure [13,14]. But still, many points remain controversial, such as the mechanism underlying CO 2 sensing; the identities of the CAs involved in this pathway; the function of CAs under low-CO 2 conditions; and the interaction of CO 2 with light, temperature, humidity, and phytohormones in influencing stomatal movement.
The primary requirement for solving these questions is the isolation of mutants. Screening dependent on thermal imaging is quite common for isolating Arabidopsis mutants with abnormal guard cell behavior. Almost all of the mutants obtained until now, including ht1, rhc1, and patrol1, were obtained using this method. Although this method has been effective in unraveling the regulation network in CO 2 -mediated stomatal movement, it is still not clear of this regulation network; thus, it is urgent to develop new screening methods.
Calcium ion (Ca 2+ ) has been shown to act as a key cellular second messenger in numerous plant processes. In Arabidopsis thaliana, abscisic acid (ABA), hydrogen peroxide, cold, and CO 2 all can stimulate cytosolic Ca 2+ ([Ca 2+ ] cyt ) oscillation, which causes stomatal closure [15]. CO 2 -induced stomatal closure is strongly Ca 2+ -dependent in Arabidopsis, consistent with previous findings in Commelina guard cells [16][17][18]. Cytosolic Ca 2+ regulates stomatal closure by two mechanisms: short-term Ca 2+ -reactive closure and long-term Ca 2+ -programmed closure [15]. Extracellular CO 2 induces changes of the [Ca 2+ ] cyt in Arabidopsis guard cells. To further dissect this signaling pathway, new components in the CO 2 response pathway that are related to the [Ca 2+ ] cyt changes need to be identified. Here, we used a novel approach for screening genetic mutants to identify proteins involved in CO 2 response. In this study, we used the Ca 2+ reporter aequorin (AEQ) to record [Ca 2+ ] cyt changes in Arabidopsis leaves in real time in order to visualize locally induced [Ca 2+ ] cyt elevations in response to CO 2 or HCO 3 − stimulus. Although this screening method had already been used for analyzing the responses of Arabidopsis to different stimuli such as salt stress, ABA, sorbitol, and cold [15], it had not been tried for screening mutants with altered stomatal responses to CO 2 or/and HCO 3 − . By using this system, we obtained several mci (mutant of HCO 3 − /CO 2 insensitive) and mcs (mutant of HCO 3 − /CO 2 sensitive) mutants. Further study with these mutants will be helpful for uncovering the mechanism for calcium-dependent CO 2 -regulated guard cell movement.

[Ca 2+ ] cys changes induced by HCO 3 − can be detected by an aequorin-based calcium imaging system
In our first experiment, we tested whether AEQ-transgenic Arabidopsis plants could be used to detect [Ca 2+ ] cys changes induced by HCO 3 − . As it is already known that the pH of incubation buffer (50 mM KCl, 0.1 mM CaCl 2 , 10 mM 2-(N-morpholino) ethanesulfonic acid (MES) and 10 μM coelenterazine) cannot be stabilized at 7.0 when the concentration of KHCO 3 is above 5 mM, a lower concentration (1 mM) that has previously been used for analyzing guard cell behavior [2] was selected to avoid the putative influence of pH.
The AEQ-transgenic Arabidopsis leaves were treated with 1 mM KHCO 3 , and after 5 min, dramatic increases in [Ca 2+ ] cys were detected in the leaves by analyzing the AEQ luminescence image (Fig. 1a, left). The average luminescence values increased from about 200 to 2300 RLU (Relative luminescence units, which represents the electrical signal values generated by stimulated photons) within 2 s of KHCO 3 addition (Fig. 1a, right). We also added incubation buffer and same concentration of KCl as controls, and found that these only caused small changes of calcium (Fig. 1a, right). These suggested that 1 mM KHCO 3 is effective for checking the cytosolic calcium changes with the AEQ system. Furthermore, when we analyzed individual guard cells within leaf epidermal strips under 1 mM KHCO 3 treatment, a visible increase in luminescence of the guard cell was found after 1 min; when luminescence values were collected continuously for about 10 min, the quantified data of luminescence images of the guard cells confirmed the visually increase in [Ca 2+ ] cys to KHCO 3 (Fig. 1b).
To further validate the AEQ-based screening method, we adopted another Ca 2+ indicator, yellow Cameleon 3.6 (YC3.6), for measuring CO 2 /HCO 3 − -induced [Ca 2+ ] cyt increases in guard cells. The YC3.6 transgenic plants showed a marked increase in [Ca 2+ ] cyt when treated with 1 mM KHCO 3 . This was consistent with the results of AEQ, suggesting that the aequorin-based system is a reliable method of [Ca 2+ ] cyt measurement (Fig. 1c).

mM HCO 3 − induces closure of Arabidopsis thaliana stomata whether in light or darkness
Since the aequorin-based system requires that samples first be incubated in incubation buffer in the dark for several hours, we measured the stomatal responses in both light and darkness. For this experiment, detached leaves of 3-week-old plants were incubated in a glass chamber with quicklime to remove CO 2 from the chamber. After 2 h of incubation in the chamber under light conditions, almost all stomata had opened very well. The opened stomata were closing after 20 min of 1 mM KHCO 3 treatment, and 60 min later, the stomatal apertures (width/length) were 0.71 ± 0.01 and 0.56 ± 0.01 for control and treated leaves, respectively (Fig. 2a).
Compared with the responses in light, the stomatal apertures were less after incubation in darkness without CO 2 . However, the stomata closed 5 min after 1 mM KHCO 3 was added, and stomatal aperture decreased to 0.20 ± 0.01 for treated leaves at 60 min; while stomatal aperture was 0.42 ± 0.01 for the untreated at this time ( Fig. 2b). The results showed that HCO 3 − -induced stomatal closure whether in light or darkness; thus, by using the aequorin-based system, it will be possible to indentify abnormal-response mutants for both stomatal movement and [Ca 2+ ] cys transient change at 1 mM KHCO 3 .

High-throughput screening for CO 2 /HCO 3 − response mutants
For high-throughput genetic screening with the aequorin-based system, we used the protocol shown schematically in Fig. 3. About 5000 AEQ-expressing Arabidopsis seeds were treated with 0.3% (w/v) ethyl methane sulfonate (EMS) and sown on soil. M 2 seeds were collected individually and screened as described in Fig. 3. The leaves of 3-week-old M 2 plants were placed in a 96-well plate and 100 µL of freshly prepared incubation buffer was added to each well for 4-6 h. AEQ luminescences of leave treatment with 1 mM KHCO 3 were then identified by using a luminescence reader (LB960, Berthold) (Fig. 3). So far, approximately 35,000 M 2 plants have been screened, and about 120 sensitive and 80 insensitive putative mutants have been identified. The selected plants were examined further for their stomatal response to KHCO 3 to narrow down the target mutants. HCO 3 − /CO 2 -induced stomatal closure of the putative mutants was assayed in M 2 and again in M 3 , 6 out of 80 putative mutants with lower luminescence showed an insensitive stomatal response to HCO 3 − /CO 2 , and 4 out of 120 putative mutants with higher luminescence displayed a hypersensitive response.

Characterization of mutants obtained by the aequorin-based screening method
By using the aequorin-based screening procedure, we identified HCO 3 − /CO 2 response mutants that appeared abnormal in both [Ca 2+ ] cys and stomatal movement. To examin the stability of the mutants obtained by this  method, two were selected for further analysis and named mci1 (insensitive response) and mcs1 (hypersensitive response). We first monitored the stomatal movement of mci1 and mcs1 in response to HCO 3 − /CO 2 .The results clearly showed that 1 mM HCO 3 − could induce stomatal closure within 30 min in mcs1 but not in wild type (Fig. 4a). For mci1, even 3 mM HCO 3 − could not induce stomatal closure after 1 h (Fig. 4b).
Consistent with the results of the screen, by comparing with wild type, AEQ luminescence intensities increased dramatically in mcs1, while no significant change was observed in mci1 in response to HCO 3 − treatment (Fig. 4c). These results further suggest that the products of MCS1 and MCI1 participate in HCO 3 − signal transduction pathways regulating both [Ca 2+ ] cys and stomatal movement.
It is necessary to make sure that only a single gene locus functions in controlling a phenotype of interest before conducting gene mapping. After crossing each of the mutants with wild type, we analyzed the segregation of the F 2 progeny. Phenotypes of F 2 plants showed   3:1 (wild-type:mci1 or mcs1) segregation, suggesting that mci1 or mcs1 was resulted from a recessive mutation. These two mutants are appropriate for subsequent gene mapping work.
Together, these data demonstrated that the highthroughput methods developed in this study are valuable for identifying new calcium-related components in the

Discussion
CO 2 influences both stomatal movement and stomatal development; however, the mechanisms of guard cell perception and transduction are still not fully clear, and the sensors that mediate CO 2 -controlled stomatal movement remain enigmatic. Previous studies have suggested that intracellular bicarbonate acts as a second messenger in guard cells involved in mediating CO 2 signal transduction [19][20][21]. To date, a number of proteins with critical roles in this signaling pathway have been identified, such as CA1 and CA4, HT1, SLAC1, RHC1, and others.
Because Ca 2+ is a key cellular second messenger, transient change in [Ca 2+ ] cyt reflects most physiology processes including CO 2 -regulated guard cell behavior. GROWTH CONTROLLED BY ABSCISIC ACID 2 (GCA2) has been proved to function downstream of both CO 2 signaling and ABA signaling by regulating [Ca 2+ ] cyt . gca2 mutant plants display decreased sensitivity of stomata to elevated CO 2 and show an abnormal [Ca 2+ ] cyt pattern in guard cells [22]. This altered pattern of [Ca 2+ ] cyt in CO 2 /HCO 3 − -treated guard cells prompted us to design a screening method to identify genes implicated in [Ca 2+ ] cyt regulation during stomatal response to CO 2 .
AEQ photoprotein has been extensively used in the Ca 2+ signaling field for almost 40 years. Because it is convenient, fast, sensitive, easy to use, and applicable to real-time measurement of [Ca 2+ ] cyt changes, we chose an AEQ-based system for our genetic screen. According to our data showing that the CO 2 /HCO 3 − -induced increase of [Ca 2+ ] cyt happened no more than 1 s after CO 2 /HCO 3 − application, almost climbed to the highest value, then dropped almost back to the baseline; the whole process only lasted for about 3 s (Fig. 1a, right), suggesting that the variation of [Ca 2+ ] cyt happened both early and rapidly in this physiological process. To identify components underlying this response, Arabidopsis mutants were usually isolated by analyzing their leaf temperature through thermal imaging. This traditional method was convenient and common, however thermal imaging takes hours to reach a steady state before detection, which may miss some important components that function earlier in the response to CO 2 .
According to a previous report about detecting stomatal responses to bicarbonate, 1 mM KHCO 3 has been used for screening [2]. As shown in Figs. 1, 2, both significantly increased bioluminescence and remarkable stomatal closure can be detected at this concentration. Before the screen, a period of dark treatment for AEQ incubation is necessary, so we conducted another preliminary test because dark can influence guard cell status. We found that even in dark treatment, 1 mM KHCO 3 still can cause closure of the stomata (Fig. 2b), further suggesting the suitability of this screening method.
By using this AEQ-based method and treatment with 1 mM KHCO 3 , we obtained mci and mcs mutants from about 35,000 M 2 seeds. We will continue to analyze these mutants and characterize the function of these genes. This series of experiments will shed light on the mechanism of calcium-mediated CO 2 /HCO 3 − response in the guard cell, which appears to occur early during CO 2 / HCO 3 − -induced stomatal closure.

Conclusions
We have developed a sensitive method for isolating stomatal CO 2 /HCO 3 − response genes that function early in the response and play a role in regulating [Ca 2+ ] cyt transient changes. This method will be helpful in elucidating the Ca 2+ -dependent regulation of stomatal response.

Plant material and growth conditions
Lines of Arabidopsis thaliana ecotype Col-0 constitutively expressing the intracellular Ca 2+ indicator AEQ (pMAQ2; a gift from Marc R. Knight) or Cameleon (YC3.6; a gift from Simon Gilroy) were used. Plants homozygous for the AEQ-transgenic Arabidopsis plant were selected from the second generation after transformation (T1 plants). One such plant, expressing a high level of AEQ, was selected for subsequent experiments.
Plants were grown in soil or in medium containing Murashige and Skoog salts (MS; PhytoTechnology Laboratories), 3% (w/v) sucrose (Sigma), and 0.6% agar (Solarbio) in controlled environmental rooms at 20 ± 2 °C. The fluency rate of white light was ~ 80-100 μmol m −2 s −1 . The photoperiod was 16 h light/8 h dark. Seeds were sown on MS medium, placed at 4 °C for 3 days in the dark, and then transferred to growth rooms.

AEQ bioluminescence-based Ca 2+ imaging
[Ca 2+ ] cys was measured using Arabidopsis plants expressing AEQ. Leaves (1 per well) were treated evenly with 150 μL of 10 μM coelenterazine (Sigma, C2230) in 96-well white culture plates 4 to 6 h before imaging and placed in the dark and in the glass chamber to remove CO 2 . AEQ bioluminescence imaging was performed using a Berthold LB985 system equipped with a light-tight box and a cryogenically cooled, back-illuminated CCD camera. The recording of luminescence was started 30 s prior to treatment and lasted for 5 min. All the treatments were carried out in the dark, and the experiments were carried out at room temperature (22-24 °C).
Similarly, guard cells were used for AEQ bioluminescence imaging. Rosette leaf epidermal peels from 3-to 4-week-old plants were placed in a microwell chamber in incubation buffer for 4-6 h in the dark. AEQ bioluminescence imaging of guard cells was performed using a bioluminescence microscope (Sclis; Biocover) equipped with a light-tight box and a cryogenically cooled, backilluminated CCD camera. The recording of luminescence was started 60 s prior to treatment and lasted for 5 min. Bright-field images were taken after AEQ imaging. All treatments were carried out in the dark, and the experiments were carried out at room temperature (22-24 °C).

Mutant screening
Arabidopsis seeds expressing AEQ were mutagenized with EMS as described previously [23]. Briefly, about 5000-10,000 seeds were imbibed overnight and then shaken in 0.3% EMS (v/v) for 15 h. The M 1 seeds were rinsed thoroughly with tap water, sterilized with 10% bleach for 30 min, and washed with sterilized water 5-8 times. M 2 seeds were harvested separately from individual M 1 plants. For screening, M 2 seeds were individually planted in soil and grown for 3 weeks. Leaves from M 2 plants were placed in a 96-well plate and 100 µL of freshly prepared incubation buffer was added to each well. Kinetic luminescence measurements were performed with an automated microplate luminescence reader (LB960; Berthold) every 0.2 s. After 3 s of luminescence counts, 100 µL of 2 mM KHCO 3 solution was automatically injected into each well to obtain a final concentration of 1 mM. Bioluminescence was recorded for 30 s per well.

Stomatal aperture bioassay
Leaves of 3-to 4-week-old seedlings were used in the stomatal aperture assays [24]. Leaves were detached before the light period started. For monitoring stomatal response to KHCO 3 in light or dark, whole leaves were incubated in stomatal buffer and then exposed to light (100 μmol m −2 s −1 ) or dark for 2 h at 25 °C in the glass chamber.
The stomatal buffer contained 50 mM KCl, 0.1 mM CaCl 2 , and 10 mM 2-(N-morpholino) ethanesulfonic acid (MES), adjusted to pH 7.0 with Tris (hydroxymethyl) aminomethane (Tris) [25,26]. Two hours later, the effects of 1 mM KHCO 3 on stomatal closure were tested. For characterization of stomatal response mutants, 3 mM KHCO 3 was also used. Prior to measuring the stomatal aperture, the adaxial epidermis and mesophyll layers were gently separated, and the epidermal strips were placed on microslides containing a drop of stomatal buffer with the desired concentration of KHCO 3 and covered with coverslips. Pictures of stomata were acquired using an inverted microscope (IX73; Olympus) at 40× magnification. Approximately 100 stomatal apertures from different leaves of each plant type were measured using Image J software (Broken Symmetry Software), and three independent experiments were performed.

Cameleon-based [Ca 2+ ] cys imaging in guard cells
The wild-type plants constitutively expressing GFP fluorescence resonance energy transfer (FRET)-based Ca 2+ sensor YC3.6, and 10 homozygous lines were generated. Rosette leaf epidermal peels from 3-week-old plants were placed in a microwell chamber in the stomatal buffer for 2 h under light (100 μmol m −2 s −1 ). Epidermal peels were treated with 1 mM KHCO 3 and ratiometric Ca 2+ imaging was performed using a confocal microscope (LSM710; Zeiss) as described previously [27]. The YC3.6 Ca 2+ sensor was excited with the 458 nm line of the argon laser. The cyan fluorescent protein (CFP; 473-505 nm) and FRET-dependent Venus (562-536 nm) emission were collected using a 458 nm primary dichroic mirror and the Meta detector of the microscope. Emission images (562-536 nm and 473-505 nm) of epidermal peels were taken, and ratiometric images before and 10 s after addition of 1 mM KHCO 3 .

Identification of MCI1 and MCS1 by MutMap analysis
We backcrossed mci1 or mcs1 to AEQ-expressing Col-0 and produced F 2 individuals. Plants with the mci1 or mcs1 phenotype were then subjected to MutMap analysis to find the mutated gene [28]. DNA of 30 F 2 progeny showing the mutant phenotype was isolated and then bulked using an equal amount of DNA from each plant. This bulked DNA was then subjected to MutMap analysis.