Brassinosteroid Involvement in Arabidopsis thaliana Stomatal Opening

Abstract

Stomata within the plant epidermis regulate CO₂ uptake for photosynthesis and water loss through transpiration. Stomatal opening in Arabidopsis thaliana is determined by various factors, including blue light as a signal and multiple phytohormones. Plasma membrane transporters, including H⁺-ATPase, K⁺ channels and anion channels in guard cells, mediate these processes, and the activities and expression levels of these components determine stomatal aperture. However, the regulatory mechanisms involved in these processes are not fully understood. In this study, we used infrared thermography to isolate a mutant defective in stomatal opening in response to light. The causative mutation was identified as an allele of the brassinosteroid (BR) biosynthetic mutant dwarf5. Guard cells from this mutant exhibited normal H⁺-ATPase activity in response to blue light, but showed reduced K⁺ accumulation and inward-rectifying K⁺ (K⁺_in) channel activity as a consequence of decreased expression of major K⁺_in channel genes. Consistent with these results, another BR biosynthetic mutant, det2-1, and a BR receptor mutant, bri1-6, exhibited reduced blue light-dependent stomatal opening. Furthermore, application of BR to the hydroponic culture medium completely restored stomatal opening in dwarf5 and det2-1 but not in bri1-6. However, application of BR to the epidermis of dwarf5 did not restore stomatal response. From these results, we conclude that endogenous BR acts in a long-term manner and is required in guard cells with the ability to open stomata in response to light, probably through regulation of K⁺_in channel activity.

Introduction

Stomatal pores surrounded by pairs of guard cells in the plant epidermis regulate gaseous exchange between plants and the atmosphere, thereby controlling photosynthesis and transpiration (Shimazaki et al. 2007). Stomata open upon exposure to light; blue light under strong red light is the most effective treatment for stomatal opening (Shimazaki et al. 2007, Inoue et al. 2010, Marten et al. 2010). A key pathway that controls stomatal opening involves the activation of plasma membrane H⁺-ATPase via blue light-receptor phototropins (Kinoshita et al. 2001, Ueno et al. 2005, Inoue et al. 2008) and its downstream inward-rectifying K⁺ (K⁺_in) channels that induce an influx of K⁺ (Schroeder et al. 1987, Brüggemann et al. 1999, Kwak et al. 2001, Lebaudy et al. 2008). The accumulation of K⁺, Cl^–, nitrate and malate^2– through the channel decreases the water potential of guard cells, causing water uptake and leading to stomatal opening (Shimazaki et al. 2007, Inoue et al. 2010). Recent studies have uncovered transcriptional regulation of the K⁺_in channels in guard cells. For example, transcription factors of ABA-RESPONSIVE KINASE SUBSTRATEs (AKSs) bind to the promoter of the K⁺_in channel KAT1 gene and directly activate its expression in guard cells (Takahashi et al. 2013). The phytohormone ABA decreases the expression of K⁺_in channel genes (Leonhardt et al. 2004) via inactivation of AKS transcription factors in guard cells, and inhibits the stomatal opening (Takahashi et al. 2013, Takahashi et al. 2016). However, it is likely that more regulatory mechanisms for transcription of the K⁺_in channel genes in guard cells remain to be discovered.

Phytohormones also regulate stomatal aperture by modulating the K⁺_in channel post-translationally. For example, ABA inhibits K⁺_in channel-mediated K⁺ uptake by suppressing channel activities (Armstrong et al. 1995, Schroeder et al. 2001, Sato et al. 2009). ABA also promotes internalization of KAT1 from the plasma membrane into endomembrane compartments by endocytosis, thereby reducing the number of K⁺_in channels functioning at the plasma membrane (Sutter et al. 2007). Auxin has been reported to activate K⁺_in channel currents (Blatt and Thiel 1994), although there are some contradictory reports of auxin action on stomatal aperture (Willmer and Fricker 1996). Brassinosteroid (BR) is another phytohormone that affects stomatal opening; its action may vary among plant species and application concentrations (Acharya and Assmann 2009, Xia et al. 2014). However, whether regulation occurs through the K⁺_in channel or by other means is unknown.

In this study, we used infrared thermography to obtain screened Arabidopsis mutants that were impaired in light-dependent stomatal opening. Consequently, we identified a new allele of the gene in the BR biosynthetic mutant dwarf5 responsible for the defect in blue light-induced stomatal opening. Together, these results indicate that endogenous BR maintains K⁺_in channel gene expression in guard cells and contributes to stomatal opening by light.

Results

Closed stomata 1 (cst1) is defective in blue light-induced stomatal opening

We obtained Arabidopsis mutants exhibiting a high leaf temperature under white light (50 µmol m^–2 s^–1) from 43,000 ethyl methanesulfonate-mutagenized glabrous1-1 (gl1) mutant plants by infrared thermography. We isolated a mutant showing closed stomata, and named it closed stomata 1 (cst1). Imaging by infrared thermography confirmed that leaf temperature in cst1 was higher than in gl1, and cst1 displayed a leaf temperature phenotype similar to that of the phot1-5 phot2-1 (phot1 phot2) double mutant in the gl1-1 background (Fig. 1A).

The stomata of gl1 are opened by blue light, whereas those of phot1 phot2 lose this response (Kinoshita et al. 2001, Doi et al. 2004, Inoue et al. 2011). Stomata in the epidermis of cst1 isolated from leaves of light-grown plants exhibited smaller apertures than those of gl1 (Fig. 1B). We prepared epidermal strips from dark-adapted leaves and measured blue light-mediated opening (Inada et al. 2004, Inoue et al. 2008, Takemiya et al. 2013). Stomata in the epidermis of gl1 barely opened in response to red light, but opened wide in response to weak blue light in the presence of red light (Fig. 1C). This blue light-dependent stomatal opening was impaired in the cst1 mutant, as has been shown previously in the phot1 phot2 double mutant. In contrast, the cst1 mutant showed clear phototropism, chloroplast movement and leaf flattening, although all of these responses are lost in the phot1 phot2 double mutant (Supplementary Fig. S1). These results indicate that phototropin-mediated stomatal opening was impaired in the cst1 mutant, while other phototropin-mediated responses were not affected.

When plants were subjected to drought stress for 14 d in light, gl1 plants wilted with chlorosis in the rosette leaves (Fig. 1D). However, cst1 did not show such strong symptoms, probably resulting from the closed stomata in this mutant (Fig. 1B).

CST1 encodes a BR biosynthetic enzyme

We performed map-based cloning using the cst1 mutant to identify the CST1 locus. Because the cst1 mutant showed a semi-dwarf phenotype (Fig. 2A), we conducted map-based cloning using the dwarf phenotype as a physiological marker. The CST1 locus was found to be present in the region between bacterial artificial chromosome clones F11F12 and F11M15 on the lower arm of chromosome 1 (Fig. 2B). This region contains DWARF5 (DWF5; AT1G50430) whose corresponding mutant is known to exhibit a semi-dwarf phenotype similar to that of cst1. DWF5 encodes a sterol Δ⁷ reductase that functions in the biosynthetic pathway for the production of the phytohormone BR. BR levels are reduced in the dwf5-2 mutant, resulting in multiple visible phenotypes including small, round, dark green rosette leaves, short stems, short pedicels and short petioles (Choe et al. 2000). Because cst1 shows phenotypes similar to these (Fig. 2A, C), we reasoned that the cst1 mutation might be an allele of dwf5. We thus determined the genomic nucleotide sequence of the DWF5 gene in the cst1 mutant and found a single nucleotide substitution of cytosine to thymine in the sixth exon (Fig. 2B). This substitution causes an amino acid substitution from Pro184 to leucine in the DWF5 protein.

We next examined whether expression of the wild-type DWF5 gene could functionally complement the cst1 mutation. A 6,345 bp genomic DWF5 fragment containing 5′- and 3′-non-coding regions was introduced into the cst1 mutant. Transformed lines in the T₃ generation exhibited completely restored growth, development (Fig. 2C) and stomatal opening in response to blue light (Fig. 2D). To confirm the involvement of DWF5 in stomatal response, we assessed the stomatal phenotype in two independent alleles, dwf5-8 (SALK_127066) and dwf5-9 (SAIL_232E05), containing a T-DNA insertion (Fig. 2B). Both dwf5-8 and dwf5-9 showed the same visible phenotypes as cst1 (i.e. dwarfism and dark green leaves) (Fig. 2E). Transcript analysis indicated that dwf5-8 is a knockdown mutant, whereas dwf5-9 appears to be a knockout mutant (Fig. 2F). Blue light-dependent stomatal opening was impaired in both mutants, as also observed in the cst1 mutant (Fig. 2G). These results indicate that the cst1 mutant is allelic to dwf5, and that DWF5 is required for blue light-dependent stomatal opening in Arabidopsis. We thus renamed the cst1 mutant dwf5-7 gl1.

Activation of plasma membrane H⁺-ATPase is not impaired in dwf5-7 gl1

The stomatal opening defect in the dwf5-7 gl1 mutant could result from an adverse effect on guard cell H⁺-ATPase. To test this, we treated epidermal strips from gl1, phot1 phot2 and dwf5-7 gl1 with the fungal toxin fusicoccin (FC), which activates guard cell H⁺-ATPase (Shimazaki et al. 1993, Kinoshita and Shimazaki 2001). In darkness, the application of FC caused large stomatal openings in the epidermis of both gl1 and phot1 phot2 but only slight opening in dwf5-7 gl1 (Fig. 3A). This result suggests that dwf5-7 gl1 is impaired in its guard cell H⁺-ATPase activity or its signaling downstream from the H⁺-ATPase. We therefore prepared guard cell protoplasts (GCPs) from gl1 and dwf5-7 gl1 rosette leaves, and examined blue light-dependent H⁺ pumping (Fig. 3B). Rates and magnitudes of pumping in GCPs from gl1 and dwf5-7 gl1 leaves were similar (Fig. 3B; Supplementary Table S1), suggesting that light-driven activation of guard cell H⁺-ATPase is not impaired in dwf5-7 gl1.

Blue light activates plasma membrane H⁺-ATPase in guard cells through phosphorylation of C-terminal threonine with subsequent binding of a 14-3-3 protein (Kinoshita and Shimazaki 1999, Ueno et al. 2005, Hayashi et al. 2011). We next measured the phosphorylation status of the H⁺-ATPase C-terminal threonine residue by protein blotting, using a 14-3-3 protein as a probe (Ueno et al. 2005, Takemiya et al. 2013). 14-3-3 binding to the H⁺-ATPase C-terminus was similar in GCPs from gl1 and dwf5-7 gl1 in response to blue light and FC application (Fig. 3C: upper panel). In addition, the amount of H⁺-ATPase in GCPs was not affected by the dwf5-7 mutation (Fig. 3C: lower panel). We further determined the phosphorylation status of H⁺-ATPase in the guard cells of intact leaves by immunohistochemistry using specific antibodies against phosphorylated threonine residues (Hayashi et al. 2011). No difference was observed between gl1 and the dwf5-7 gl1 mutants tested (Supplementary Fig. S2). Together, these results demonstrate that the expression and activation of guard cell H⁺-ATPase are not affected in dwf5-7 gl1, suggesting that the signaling pathway for stomatal opening in this mutant is impaired downstream of H⁺-ATPase.

K⁺ accumulation and K⁺_in channel activity in dwf5-7 gl1 guard cells

Given these findings, we reasoned that guard cell accumulation of K⁺ following H⁺-ATPase activation may be altered in dwf5-7 gl1. Cell extracts were obtained from illuminated GCPs and subjected to HPLC to determine the amount of K⁺. K⁺ levels increased by a factor of 1.84 following blue light excitation in GCPs isolated from gl1 (Fig. 4A). The increment in K⁺ accumulation was decreased by 45% (1.47-fold) in GCPs isolated from dwf5-7 gl1. We next determined guard cell K⁺_in channel activity by monitoring K⁺ uptake under a whole-cell patch clamp configuration. Steady-state current–voltage curves showed that dwf5-7 gl1 guard cells yielded a lower K⁺ uptake amplitude than did those of gl1 (Fig. 4B). These results indicate that dwf5-7 gl1 guard cells showed reduced K⁺ accumulation and that this phenotype most probably results from a reduction in K⁺_in channel activity.

Multiple members of K⁺_in channels, such as KAT1, KAT2, AKT1, AKT2/3, AKT5 and AtKC1, are expressed in Arabidopsis guard cells and function redundantly in stomatal opening (Szyroki et al. 2001, Véry and Sentenac 2003, Gambale and Uozumi 2006, Harada and Shimazaki 2009, Takahashi et al. 2013). Among these, we determined the mRNA levels of KAT1, KAT2 and AKT1 in gl1 and dwf5-7 gl1 GCPs by quantitative reverse transcription–PCR (qRT–PCR). In all of these K⁺_in channel genes, expression was decreased by approximately 40% in dwf5-7 gl1 GCPs compared with those in gl1 (Fig. 4C). The decreased expression of KAT1 in dwf5-7 gl1 was also consistent with GUS (β-glucuronidase) expression in transgenic Arabidopsis expressing KAT1pro:GUS in the dwf5-7 gl1 background (Fig. 4D).

Since the basic helix–loop–helix (bHLH) transcription factor AKS1 is known to enhance the expression of K_in channels through KAT1 gene activation in guard cells (Takahashi et al. 2013), we reasoned that AKS1 expression may be reduced in dwf5-7 gl1 guard cells. However, AKS1 transcript levels were unaltered in the dwf5-7 gl1 mutant (Fig. 4C). These results suggest that BR is needed for K⁺_in channel expression, independent of AKS1 in guard cells.

The decreased expression of K⁺_in channels is probably responsible for K⁺_in current reduction and K⁺ accumulation in dwf5-7 gl1 guard cells. To verify this, we expressed KAT1 or AKT1 cDNA in dwf5-7 gl1 under the control of the guard cell-specific promoter GC1 (Yang et al. 2008, Hu et al. 2010; Supplementary Fig. S3A). In response to blue light, stomata opened more widely in both of these transgenic lines compared with those of dwf5-7 gl1 (Supplementary Fig. S3B). However, stomatal apertures were significantly smaller in these transgenic lines than in gl1. These results indicate that reduced stomatal opening in dwf5-7 gl1 is at least partially due to the decreased expression of K⁺_in channels.

BR is required for stomatal opening

If BR is required for stomatal opening, we can expect that other BR-related mutants will exhibit a stomatal opening phenotype similar to that of dwf5-7 gl1. We therefore measured stomatal apertures using a BR biosynthetic mutant, de-etiolated 2-1 (det2-1), and a BR receptor mutant, brassinosteroid insensitive1-6 (bri1-6). Both det2-1 and bri1-6 exhibited dwarfism under our growth conditions, as did dwf5-7 gl1 (Fig. 5A). Stomata in gl1, En-2 and Col leaves showed large apertures (>3 m) but apertures in the dwf5-7 gl1, det2-1 and bri1-6 mutants were small, like those of phot1 phot2 double mutants under light (Fig. 5B).

Next, we examined blue light-dependent and FC-induced stomatal opening in epidermal strips isolated from leaves of the mutants. Stomata of dwf5-7 gl1, det2-1 and bri1-6 showed decreased stomatal aperture in response to both blue light (Fig. 5C; Supplementary Fig. S4) and FC (Fig. 5D). Among these BR-related mutants, this impairment was strongest in bri1-6 (Fig. 5B–D). Similarly, bri1-6 exhibited strong resistance to drought stress (Supplementary Fig. S5). These results, showing that all of the BR biosynthesis and BR receptor mutants examined exhibited impaired stomatal opening in response to blue light, suggest that BR is required for stomatal opening.

We observed that the closed-stomatal phenotypes among the BR-related mutants are derived from a decrease in the response of stomatal opening, because the stomatal sizes of all BR-related mutants tested were not significantly different from those of the corresponding control plants (Supplementary Fig. S6).

To investigate the effect of BR on the light response of stomata, we treated epidermises from gl1 and dwf5-7 gl1 with a specific BR brassinolide for 4 h and measured blue light-dependent stomatal openings (Supplementary Fig. S7). Following this short-term treatment, the light response in dwf5-7 gl1 stomata was not recovered, and stomatal openings in the gl1 epidermis were not affected. Thus, we next grew plants of gl1, dwf5-7 gl1, Col, det2-1, En-2 and bri1-6 on the hydroponic culture medium containing 0.1 µM brassinolide for 3 weeks (Supplementary Fig. S8) and measured blue light-dependent stomatal openings (Fig. 6). This long-term application of BR partially recovered dwarfism and completely recovered stomatal opening in dwf5-7 gl1 and det2-1. In contrast, the effects of BR application were not observed in bri1-6. From these findings, we conclude that BR does not directly promote stomatal opening and that BR action related to stomatal opening may require a longer term effect of endogenous BR in Arabidopsis thaliana.

Discussion

Effect of BR on guard cell K⁺_in channel expression and stomatal opening

In stomatal opening, blue light activates plasma membrane H⁺-ATPase via phototropin-dependent phosphorylation in guard cells (Kinoshita et al. 2001, Ueno et al. 2005, Hayashi et al. 2011). Activated H⁺-ATPase increases the hyperpolarization of the guard cell plasma membrane (Assmann et al. 1985, Shimazaki et al. 1986), and this hyperpolarization drives stomatal opening through K⁺ uptake via guard cell K⁺_in channels (Schroeder et al. 2001, Shimazaki et al. 2007, Marten et al. 2010). The BR biosynthetic dwf5 mutant guard cells exhibited impairment in blue light-induced stomatal opening and reduced K⁺ accumulation and K⁺_in channel activity (Figs. 2D, G, 4A, B). In dwf5-7 gl1 guard cells, K⁺_in channel gene expression was reduced (Fig. 4C, D). In contrast, guard cell H⁺-ATPase was normally activated by blue light in dwf5-7 gl1 and bri1-6 mutant guard cells (Fig. 3B; Supplementary Fig. S2B). Furthermore, all of the BR-related mutants tested showed small stomatal apertures in response to blue light (Fig. 5B, C). We conclude that BR has a positive role in stomatal opening. BR is perceived in the extracellular domain of the BRI1 receptor on the cell surface (He et al. 2000, Wang et al. 2001). Activated BRI1 then initiates intracellular signaling via sequential components, and finally regulates the expression of BR-responsive genes (Belkhadir and Chory 2006, Clouse 2011). These findings suggest that BR may be synthesized in cells other than guard cells and acts via BRI1 to render normal stomatal function.

aperture is precisely controlled by phytohormones to adapt plants to their ever-changing environments. K⁺_in channels play a crucial role in light-induced stomatal opening; ABA inhibits the opening response by suppressing K⁺_in channel activity (Armstrong et al. 1995, Assmann and Shimazaki 1999, Sutter et al. 2007, Takahashi et al. 2013). These findings, combined with those of the present study, indicate that ABA acts in balance with BR in controlling the expression of K⁺_in channels. Since AKS1 directly promotes KAT1 gene expression in guard cells and contributes to blue light-induced stomatal opening (Takahashi et al. 2013), we expected that AKS1 levels might be reduced in guard cells in the absence of BR. However, AKS1 expression was not affected in dwf5-7 guard cells (Fig. 4C), suggesting that BR-mediated K⁺_in channel gene expression is independent of AKS1. The identification of BR-inducible transcription factors for K⁺_in channel genes in guard cells will therefore be an important challenge in future research.

note that the expression of KAT1 or AKT1 in guard cells did not completely restore blue light-dependent stomatal opening in the dwf5-7 gl1 mutant (Supplementary Fig. S3B). In addition, stomatal opening was barely induced in dwf5-7 gl1, although K⁺ accumulation and K⁺_in channel activity were inhibited but partially retained in dwf5-7 gl1 guard cells (Figs. 1C, 4B, C). Therefore, it is likely that the dwf5-7 gl1 stomatal phenotype is also derived from factors other than K⁺_in channels. BR may enhance the expression of other components required for stomatal opening.

BR roles in stomatal movement

BR-regulated cell phenotypes are brought about by a number of distinct processes, including ion transport, microtubule reorganization, cell wall organization and extensibility through the expression of many genes in Arabidopsis (Zurek et al. 1994, Thummel and Chory 2002, Zhang et al. 2005, Sun et al. 2010). All of these processes may be involved in both cell expansion and stomatal opening mediated by guard cells. However, guard cell sizes in the BR-related mutants tested were similar to those in control plants (Supplementary Fig. S6). It is possible that the sizes of guard cells and other cells are regulated by BR in different ways. Our results indicate that endogenous BR mediates sufficient K⁺ uptake via K⁺_in channel expression in guard cells. However, evidence that BR enhances K⁺_in channel expression in regions other than stomatal guard cells is still lacking from transcriptome analyses of wild-type seedlings or BR signaling and BR biosynthetic mutants (Goda et al. 2002, Müssing et al. 2002, Sun et al. 2010). BR-regulated K⁺_in channel expression may therefore be a response that is specific to guard cells.

BR involvement in stomatal movement has been reported in several plants to date (Acharya and Assmann 2009). For example, application of BR to the epidermis of Vicia faba promotes stomatal closure and inhibits stomatal opening (Haubrick and Assmann 2006). In contrast, pre-treatment of Pinus banksiana seedlings with homobrassinolide was found to inhibit stomatal closure under water stress conditions (Rajasekaran and Blake 1999). The effect of exogenous BR on stomatal response was found to depend on the concentration of 24-epibrassinolide (EBR) in tomato; low EBR concentrations induced stomatal opening, whereas high concentrations induced stomatal closure (Xia et al. 2014). Recently, EBR application on Arabidopsis leaves was also shown to induce stomatal closure (Shi et al. 2015, Ha et al. 2016). However, the BR biosynthetic Arabidopsis mutants sax1 and bul1-1/dwf7-3/ste1-4 have been shown to exhibit smaller stomatal apertures than the wild type (Ephritikhine et al. 1999a, Ephritikhine et al. 1999b, Catterou et al. 2001). Thus, the effects of BR application on stomatal movement may vary between plant species, although endogenous BR is likely to act positively on stomatal opening (Xia et al. 2014). In our study, short-term application of BR to the epidermis did not show any effect on stomatal aperture, but long-term application of BR completely recovered stomatal opening in dwf5-7 gl1 and det2-1 under the blue light conditions (Fig. 6; Supplementary Fig. S7). We conclude that endogenous BR may act in a long-term manner and confer the ability to open stomata during functional maturation of guard cells in A. thaliana.

Although both DWF5 and DWF7 catalyze early steps in the BR biosynthetic pathway (Choe et al. 2000, Catterou et al. 2001), the bul1-1/dwf7-3/ste1-4 mutant is also characterized as a sterol-deficient mutant. Because DWF5 catalyzes the biosynthetic step immediately downstream of DWF7, the dwf5 gl1 mutant may also be a sterol-deficient mutant. However, defects of stomatal opening in both mutants were recovered by application of BR (Catterou et al. 2001; Fig. 6). In addition, det2-1 and bri1-6 mutants which are not sterol-deficient mutants also exhibited the impairment of stomatal opening like the dwf5 gl1 mutant (Fig. 5B, C). These results indicate that stomatal opening requires BR rather than sterol in A. thaliana plants.

Roles of BR-regulated stomatal opening in plant growth

BR-regulated processes enable plants to grow optimally and adapt to various environmental conditions (Belkhadir and Chory 2006). The typical phenotype of BR synthetic and BR signaling mutants is an extreme growth defect and dwarfism (Fig. 2A; Clouse and Sasse 1998, Choe et al. 2000, Friedrichsen and Chory 2001), but the cause of these phenotypes is not fully understood at the physiological level. In this study, we demonstrated that both BR synthetic and BR signaling mutants exhibit severe impairment in stomatal opening in response to light (Fig. 5B, C). Open stomata enhance CO₂ uptake for photosynthesis and generate a transpirational stream that facilitates the uptake of mineral nutrients into the roots and their distribution in various plant tissues that are essential for plant growth (Shimazaki et al. 2007). Stomata have a crucial role in regulating upward elongation of the plant body. Indeed, a rare fern lacking stomata has been shown to exhibit limited growth and low stature of the aerial parts (Keeley et al. 1984). Thus, the growth defect in BR-related mutants may be at least partially derived from lesions associated with defects in the stomatal opening. In support of this, overexpression of the BR biosynthetic gene DWARF4 or the BR receptor gene BRI1 enhances photosynthesis and growth in Arabidopsis (Choe et al. 2001, Oh et al. 2011).

Materials and Methods

Plant materials

Wild-type En-2 and mutants of dwf5-8, dwf5-9, bri1-6 and det2-1 were obtained from the Arabidopsis Biological Resource Center (Ohio State University). gl1 (gl1-1), phot1 phot2 (gl1-1 phot1-5 phot2-1), dwf5s and det2-1 are in the Columbia (Col) ecotype background, and bri1-6 is in the En-2 ecotype background. gl1 was used as a control of phot1 phot2 and exhibits similar phototropin-mediated responses to Col (Inoue et al. 2011). Ethyl methanesulfonate-mutagenized M₂ seeds were purchased from Lehle Seeds (gl1 background). Plants were grown in soil under a 14 h white light (50 µmol m^–2 s^–1) and 10 h dark cycle at 20–25°C at a relative humidity of 55–75% in a temperature-controlled growth room.

Hydroponic culture growth of A. thaliana plants

To observe a long-term effect of exogenous BR on stomatal opening, the BR-related mutant plants were grown under BR application. Because A. thaliana plants grown on agarose medium in airtight conditions might not show normal stomatal responses, we adopted a hydroponic culture system (Norén et al. 2004). Plants were grown on half-strength Murashige and Skoog (MS) medium for 10 d. Then, the plants were transferred onto the hydroponic culture with or without brassinolide at 0.1 µM and further grown for 3 weeks under the conditions described.

Thermal imaging

Leaf temperature was measured using an infrared thermograph (TVS-8500; NEC Avio Infrared Technology). Plants were grown for 3 weeks under the conditions described above, and thermal images were taken 4 h after the start of the light period. Thermal images were analyzed using PE Professional software (NEC Avio Infrared Technology).

Stomatal aperture

Stomatal aperture in the abaxial epidermis was determined by microscopic examination (Inoue et al. 2008). To measure stomatal aperture during growth, epidermal fragments were prepared 3 h after the start of the light period, and the stomatal aperture was immediately measured (Figs. 1B, 5B). To measure stomatal opening in response to blue light, FC or BR, epidermal fragments were isolated from the dark-adapted plants and treated with light, FC or brassinolide (Inoue et al. 2008, Takahashi et al. 2013, Takemiya et al. 2013) (Figs. 1C, 2D, G, 3A, 5C, D, 6; Supplementary Figs. S3B, S7).

Preparation of guard cell protoplasts

GCPs from Arabidopsis leaves were prepared enzymatically (Ueno et al. 2005), with slight modifications: we used Macerozyme R-10 (Yakult Pharmaceutical Industry) instead of pectolyase Y-23 (Seishin Pharmaceutical). The purity of the GCPs was higher than 97% on a cellular basis.

Measurement of H⁺ pumping in Arabidopsis guard cell protoplasts

Blue light-dependent H⁺ pumping by GCPs was measured using a glass pH electrode (Ueno et al. 2005, Inoue et al. 2008). We used 100 µg of protein of GCPs per measurement.

Immunoblot and protein blot

GCPs were incubated in a glass cuvette containing 10 mM MES-KOH (pH 6.0), 1 mM CaCl₂, 0.4 M mannitol and 10 mM KCl at 24°C, and continuously pre-illuminated with red light (600 µmol m^–2 s^–1) for 30 min; a blue light pulse (100 µmol m^–2 s^–1, 30 s) was then superimposed on the red light. The reaction with blue light was terminated 2.5 min after the pulse was started (Ueno et al. 2005). An aliquot of the GCP suspension (100 µg of protein) was centrifuged at 10,000 × g for 15 s, and the pellet was mixed with 50 µl of a solution containing 10 mM MOPS-KOH (pH 7.5), 2.5 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 20 µM leupeptin and 0.4% (w/v) Triton X-100 to suspend the pellet. The suspension was solubilized by adding 25 µl of SDS sample buffer containing 3% SDS, 30% (w/v) sucrose, 10% (v/v) 2-mercaptoethanol, 0.012% (w/v) Coomassie Brilliant Blue, 1 mM EDTA and 30 mM Tris–HCl (pH 8.0) at room temperature. The solubilized proteins were subjected to SDS–PAGE. Immunoblot and protein blot H⁺-ATPase analyses were carried out using 14-3-3 protein (Arabidopsis GF14φ as described by Kinoshita et al. (2003).

Immunostaining

Immunohistochemical determination of plasma membrane H⁺-ATPase in leaf epidermis guard cells was performed as described by Hayashi et al. (2011). Since the bri1-6 mutant showed the most severe dwarf phenotype among the mutants tested, we could not obtain sufficient rosette leaves for GCP preparation. We therefore performed immunostaining to determine the phosphorylation status of guard cell H⁺-ATPase in intact leaves instead of a protein blot using GCPs.

Measurement of K⁺ accumulation in guard cells

Blue light-dependent K⁺ accumulation in GCPs was determined using a cation-exchange chromatography system (JASCO) as described by Takahashi et al. (2013).

Patch clamp analysis

Arabidopsis guard cell protoplasts were enzymatically isolated from rosette leaves of 4- to 6-week-old plants using a digestion solution containing 1% (w/v) Cellulase R-10, 0.5% (w/v) macerozyme R-10, 0.5% (w/v) bovine serum albumin, 10 mM ascorbic acid, 0.1 mM KCl, 0.1 mM CaCl₂ and 0.5 M d-mannitol (pH 5.5) as described by Pei et al. (1997). Whole-cell patch clamp recordings (I_Kin) were performed (Ichida et al. 1997, Takahashi et al. 2013). The patch clamp pipette solution contained 30 mM KCl, 70 mM K-Glu, 2 mM MgCl₂, 6.7 mM EGTA, 3.35 mM CaCl₂, 5 mM ATP and 10 mM HEPES-Tris (pH 7.1). The patch clamp bath solution contained 30 mM KCl, 40 mM CaCl₂, 2 mM MgCl₂ and 10 mM MES-Tris (pH 5.5). Osmolarities of the pipette and bath solutions were adjusted to 500 and 485 mmol kg^–1, respectively, with d-sorbitol. Leak currents were not subtracted.

Generation of transgenic plants

To complement the dwf5-7 gl1 mutant phenotype, a 6,345 bp genomic fragment of the DWF5 gene that contains the 3′-untranslated region and the coding and promoter regions was amplified from genomic DNA using oligonucleotide primers 5′-GGGGTACCGTTGTGAGTGACACGAAAACAGTG-3′ and 5′-GGGGTACCCCATTGACCAGAGAGATTCTACAG-3′. The PCR product was cloned into the pCAMBIA1300 KpnI site, and the resulting construct was introduced into dwf5-7 gl1 using Agrobacterium tumefaciens GV3101.

For the GUS assay, a 3,779 bp genomic fragment containing the promoter region of the KAT1 gene was amplified from genomic DNA using primers 5′-GGATCCTCCTTACGATTTTGACC-3′ and 5′-CCAAGAGATCGACATCTTTTTGATGATC-3′. The PCR product was subcloned into the pCR8-TOPO vector (Invitrogen, Thermo Fisher Scientific), and the genomic DNA region was further inserted into the GUS expression vector pGWB433 using the Gateway system (Invitrogen). The resulting vector was introduced into dwf5-7 gl1 and gl1 plants using A. tumefaciens GV3101. The obtained KAT1pro:GUS/ gl1 line was crossed with dwf5-7 gl1 to generate the KAT1pro:GUS/ dwf5-7 gl1 line.

To generate GC1pro:KAT1 and GC1pro:AKT1 vectors, coding regions of KAT1 and AKT1 cDNAs were amplified by RT–PCR using primers 5′-GCCTCTAGAAAGATGTCGATCTCTTGGACTCG-3′ and 5′-GCCTCTAGATCAATTTGATGAAAAATACAAATGATCACC-3′ for KAT1, and 5′-GCCTCTAGAGTGATGAGAGGAGGGGCTTTGTTATGC-3′ and 5′-GCCTCTAGATTAAGAATCAGTTGCAAAGATGAGATGATC-3′ for AKT1. The products were inserted into the XbaI site of the pPZP211-GC1pro vector (Kinoshita et al. 2011). The resulting vector was introduced into dwf5-7 gl1 using A. tumefaciens GV3101.

RT–PCR and qRT–PCR analyses

Total RNA was extracted from rosette leaves and GCPs using the RNeasy Plant Mini Kit (Qiagen) according to the manufacturer’s instructions. First-strand cDNA was synthesized from 0.5 µg of total RNA using the PrimeScript II 1st strand cDNA Synthesis Kit using oligo(dT)_12–18 primers (TAKARA). PCRs were performed using primers 5′-ATGGCGGAGACTGTACATTCTCC-3′ and 5′-TCAATAAATTCCCGGAATGATCCTG-3′ for DWF5, 5′-ACTTTACGCCAGTGGTCGTACAAC-3′ and 5′-AAGGACTTCTGGGCACCTGAATCT-3′ for ACT8, 5′-ATGTCGATCTCTTGGACTCG-3′ and 5′-TCAATTTGATGAAAAATACAAATGATCACC-3′ for KAT1, and 5′-ATGAGAGGAGGGGCTTTGTTATGC-3′ and 5′-TTAAGAATCAGTTGCAAAGATGAGATGATC-3′ for AKT1.

qRT–PCR was performed using Brilliant II SYBR Green QPCR master mix and an MX-3000 system (Stratagene). PCR was performed using the specific primers 5′-TCCGACACTGCTCTAATGGATG-3′ and 5′-TCTCTTTTCTTCTCGTTTGCATGGC-3′ for KAT1, 5′-TCCGTCCTGCATCAAGAGAAGTG-3′ and 5′-TACCAGCTTCCCGGCTATGTC-3′ for AKT1, 5′-ACGCTAGTGATCAAGGACATGG-3′ and 5′-AACTCTTTTCGTTTCGCCATAACCAG-3′ for KAT2, and 5′-CTCCACGATCATCAACACCAG-3′ and 5′-TCAACGTTGTCGCTTGTGTCG-3′ for AKS1. Quantification was performed using a comparative cycle threshold method, and the relative amount of PCR product for each gene was normalized to the ACT8 PCR product as an internal control amplified using the above primers.

Supplementary data

Supplementary data are available at PCP online.

Funding

This work was supported by the Ministry of Education, Sports, Science, Culture, and Technology of Japan [a Grant-in-Aid for Scientific Research on Priority Areas (No. 17084005), a Grant-in-Aid for Scientific Research (No. 21227001) to K.S., a Grant-in-Aid for the Japan Society for the Promotion of Science (JSPS) Fellows (No. 10J00254) and a Grant-in-Aid for Young Scientists (B) (No. 25840105) to S.I.].

Abbreviations

Abbreviations
 
• BR

  brassinosteroid

 
• DWF5

  DWARF5

 
• FC

  fusicoccin

 
• GUS

  β-glucuronidase

 
• qRT–PCR

  quantitative reverse transcription–PCR

 
• RT–PCR

  reverse transcription–PCR

Acknowledgments

We would like to thank Professor G. Goshima for a constructive review of the manuscript. We thank the Arabidopsis Biological Resource Center (ABRC) for providing seed stocks, and M. Inoue-Aibe, N. Nishihara-Seki and N. Ono in our laboratory for their helpful technical assistance.

Disclosures

The authors have no conflicts of interest to declare.

References