A plastid protein crucial for Ca2+-regulated stomatal responses


  • Stefan Weinl,

    1. Molekulare Entwicklungsbiologie der Pflanzen, Institut für Botanik und Botanischer Garten, Universität Münster, Schlossplatz 4, 48149 Münster, Germany;
    Search for more papers by this author
  • Katrin Held,

    1. Molekulare Entwicklungsbiologie der Pflanzen, Institut für Botanik und Botanischer Garten, Universität Münster, Schlossplatz 4, 48149 Münster, Germany;
    Search for more papers by this author
  • Kathrin Schlücking,

    1. Molekulare Entwicklungsbiologie der Pflanzen, Institut für Botanik und Botanischer Garten, Universität Münster, Schlossplatz 4, 48149 Münster, Germany;
    Search for more papers by this author
  • Leonie Steinhorst,

    1. Molekulare Entwicklungsbiologie der Pflanzen, Institut für Botanik und Botanischer Garten, Universität Münster, Schlossplatz 4, 48149 Münster, Germany;
    Search for more papers by this author
  • Sebastian Kuhlgert,

    1. Institut für Biochemie und Biotechnologie der Pflanzen, Universität Münster, Hindenburgplatz 55, 48143 Münster, Germany
    Search for more papers by this author
  • Michael Hippler,

    1. Institut für Biochemie und Biotechnologie der Pflanzen, Universität Münster, Hindenburgplatz 55, 48143 Münster, Germany
    Search for more papers by this author
  • Jörg Kudla

    1. Molekulare Entwicklungsbiologie der Pflanzen, Institut für Botanik und Botanischer Garten, Universität Münster, Schlossplatz 4, 48149 Münster, Germany;
    Search for more papers by this author

Author for correspondence:
Stefan Weinl
Tel: +49 251/83 24811
Fax: +49 251/83 23311
Email: Stefan.Weinl@uni-muenster.de


  • • Guard cell movements are regulated by environmental cues including, for example, elevations in extracellular Ca2+ concentration. Here, the subcellular localization and physiological function of the Ca2+-sensing receptor (CAS) protein was investigated.
  • • CAS protein localization was ascertained by microscopic analyses of green fluorescent protein (GFP) fusion proteins and biochemical fractionation assays. Comparative guard cell movement investigations were performed in wild-type and cas loss-of-function mutant lines of Arabidopsis thaliana. Cytoplasmic Ca2+ dynamics were addressed in plants expressing the yellow cameleon reporter protein YC3.6.
  • • This study identified CAS as a chloroplast-localized protein that is crucial for proper stomatal regulation in response to elevations of external Ca2+. CAS fulfils this role through modulation of the cytoplasmic Ca2+ concentration.
  • • This work reveals a novel role of the chloroplast in cellular Ca2+ signal transduction.


Stomata are pores on the surface of plants which regulate the uptake of carbon dioxide for photosynthesis and the loss of water vapor by transpiration (Hetherington & Woodward, 2003). The aperture of the pore is controlled by the state of turgor of the two guard cells. Many environmental signals (e.g. humidity, light and CO2) and endogenous plant hormones such as abscisic acid (ABA) and auxin modulate stomatal aperture (Assmann & Wang, 2001; Hetherington, 2001; Schroeder et al., 2001; Outlaw, 2003; Vavasseur & Raghavendra, 2005; Roelfsema & Hedrich, 2005). A plethora of studies have established that guard cell signaling is controlled by a network of components influencing turgor, gene expression, cytoskeletal dynamics, protein trafficking and ion channel activity (Hetherington & Woodward, 2003). An intricate signaling network interconnecting second messengers such as cyclic ADP ribose (cADPR), inositol-3-phosphate (IP3) and sphingosine-1-phosphate (SP1P) with alterations in the concentrations of plant hormones such as ABA and with the regulation of ion channels has been shown to control guard cell aperture in response to environmental cues. Cell physiological investigations have revealed that an increase in cytoplasmic Ca2+ concentration ([Ca2+]cyt) represents a key component involved in coupling a range of extracellular signals such as CO2, ABA, CaCl2 and H2O2 to changes in stomatal aperture.

The function of Ca2+ as a secondary messenger is not restricted to guard cells. Temporally and spatially defined modulations of cellular Ca2+ concentration represent primary events in response to many signals and also regulate developmental programs and various physiological processes (Hetherington & Brownlee, 2004). Such signals include plant hormones, light, stress factors, and pathogenic or symbiotic elicitors (Trewavas & Knight, 1994; Ehrhardt et al., 1996; Knight et al., 1996, 1997; Neuhaus et al., 1997; Sanders et al., 1999, 2002; Harper & Harmon, 2005). In addition, physiological processes such as guard cell regulation, root hair elongation and pollen tube growth are accompanied by distinct spatio-temporal changes in calcium concentration (Evans et al., 2001; Sanders et al., 2002).

Recent studies suggest that a Ca2+ signal is not only represented by the Ca2+ concentration but also by spatial and temporal information, including Ca2+ localization and oscillation (Allen et al., 2000; Allen et al., 2001; Young et al., 2006). Critical for the spatially distinct generation of cellular calcium transients are the stores from which this messenger ion is released. Calcium signals are generated through the opening of channels that either mediate Ca2+ entry from outside the cell or mediate release from intracellular stores. The vacuole and the endoplasmatic reticulum represent major intracellular Ca2+ sinks with endogenous Ca2+ concentrations in the millimolar range, while the resting concentration of Ca2+ in the cytoplasm has been estimated to be approx. 200 µM (Sanders et al., 1999). In addition, plant nuclei represent a potential calcium store and stimulus-induced dynamic changes in nuclear calcium concentration that appear to be independent of cytoplasmic calcium signaling have been reported (Pauly et al., 2000). Although chloroplasts have been shown to contain high concentrations of Ca2+ (between 4 and 23 mM; Portis & Heldt, 1976) their potential role in cellular Ca2+ homeostasis and signaling has remained largely unexplored (Johnson et al., 2006). Light stimulates uptake of Ca2+ into the chloroplast but the Ca2+ concentrations in the stroma do not change significantly during illumination (Kreimer et al., 1988). However, circadian chloroplast Ca2+ oscillations were observed after transfer of plants to constant darkness (Johnson et al., 1995). Such dark-induced increases in stroma Ca2+ concentrations precede the generation of elevations of [Ca2+]cyt in tobacco (Nicotiana plumbaginifolia) leaves (Sai & Johnson, 2002).

In addition, several recent studies point to an unexpected functional interconnection of the chloroplast-localized Ca2+ pool with cytoplasmic signaling events. Characterization of the chloroplast membrane protein Pisum post-floral-specific gene 1 (PPF1) revealed that it represents a putative calcium ion carrier that affects the flowering time of transgenic Arabidopsis thaliana by modulating Ca2+ storage capacities in chloroplasts (Wang et al., 2003; Li et al., 2004). Interaction of Lotus japonicus with mycorrhiza-forming fungi critically involves the function of the two loci CASTOR and POLLUX (Ehrhardt et al., 1996). Both proteins are indispensable for microbial admission to plant cells and for the occurrence of intracellular calcium spiking in response to symbiotic stimulation (Imaizumi-Anraku et al., 2005). Surprisingly, positional cloning of both genes revealed that they encode chloroplast-localized proteins with potential ion channel function, thereby assigning a novel function to chloroplasts in generating cytoplasmic Ca2+ signatures (Imaizumi-Anraku et al., 2005).

Here we describe the functional analysis of the plant-specific protein Ca2+-sensing receptor (CAS). This 42-kDa Ca2+-binding protein was originally described as a plasma membrane-localized extracellular Ca2+-sensing protein exhibiting low-affinity/high-capacity Ca2+ binding through an N-terminal domain of the protein (Han et al., 2003). Functional characterization of CAS antisense lines indicated a function of the protein in regulating stomatal responses to elevation of extracellular Ca2+ concentrations and indicated that CAS mediates Ca2+-induced [Ca2+]cyt increases.

In contrast to the proposed plasma membrane localization, comprehensive proteomics-based studies identified the CAS protein in A. thaliana as well as in the green alga Chlamydomonas reinhardtii in the chloroplast thylakoid membrane (Friso et al., 2004; Allmer et al., 2006). In this study we verified the chloroplast localization of CAS by subcellular fractionation analyses and investigation of green fluorescent protein (GFP) fusion proteins in protoplasts and leaf guard cells. Analysis of two independent T-DNA induced loss-of-function alleles uncovered a crucial function of CAS in regulating stomatal closure in response to elevations of extracellular Ca2+. This mutant phenotype coincided with impaired increases in the concentration of cytosolic Ca2+ normally induced in response to this stimulus. Consequently, our study reveals a crucial function of chloroplasts and in particular of the thylakoid-localized CAS protein in regulating guard cell responses and uncovers a critical contribution of these organelles to the generation and fine-tuning of cytoplasmic calcium signals.

Materials and Methods

General methods, gene expression analysis and plant material

Molecular biology methods were applied according to standard procedures (Sambrook & Russell, 2001). A list of primers used in this work can be obtained upon request. The identity of all plasmid constructs generated in this study was verified by sequencing. Gene expression analysis of CAS by quantitative RT-PCR and RT-PCR investigation of CAS gene expression in the T-DNA lines were performed as described previously (D’Angelo et al., 2006). Soil-grown Arabidopsis thaliana (L.) Heynh. cv. Columbia-0 plants were used in this work and cultivated in a growth chamber under a 16 h light : 8 h dark cycle with 70% atmospheric humidity and a 22 : 18°C (day : night) temperature regime. T-DNA insertion lines were obtained from NASC (SALK_070416) and GABI-KAT (GABI_665G12) stockcenters, respectively. The T-DNA insertion positions in the gene were confirmed by PCR using T-DNA border primers and CAS gene specific primers followed by DNA sequencing of the PCR products for both flanking sides. Co-segregation analyses of the T-DNA insertion with the kanamycin (cas-1) and sulfadiazine (cas-2) marker genes of the T-DNAs in segregating F2 populations revealed a single T-DNA insertion in both mutant lines.

Database and phylogenetic analyses

CAS-related proteins from other species were identified by comparisons of the A. thaliana CAS protein sequence with entries in the GenBank database (http://www.ncbi.nlm.nih.gov) using the blastp algorithm (Altschul et al., 1990). Deduced protein sequences were aligned with ClustalW (Thompson et al., 1994). The resulting multiple alignment was processed with bioedit (Hall, 1999) to determine the homology scores. To create a phylogenetic tree, amino acid sequences were aligned together with a representative from the related rhodanese-like protein family from Arabidopsis as an out-group. Tree reconstruction based on the neighbor-joining method and confirmation of tree topology by bootstrap analysis (1000 replicates) were performed with mega4 (Tamura et al., 2007).

Isolation of intact chloroplasts and thylakoid membranes

Intact chloroplasts and thylakoid membranes were isolated from transiently transformed tobacco (Nicotiana benthamiana (Domin)) leaves as described in Ruf et al. (2000). Protein analyses by SDS-PAGE, western blot and immunodetection were performed as published previously (Hippler et al., 2001). The CAS::StrepII fusion protein was detected using the Strep-Tactin horseradish peroxidase (HRP) conjugate (IBA, Göttingen, Germany) against the StrepII-tag.

Transient transformation of tobacco leaves and fluorescence microscopy

Infiltration of N. benthamiana leaves was performed as described previously (Walter et al., 2004). All constructs for localization studies were cloned into the pGPTV-II-GFP5 vector containing either 35S or Mannopine synthase (MAS) promoters (Walter et al., 2004). Plasmid introduction into Agrobacterium tumefaciens (GV3101 pMP90) and infiltration of N. benthamiana leaves were performed as described by Walter et al. (2004). Protoplast preparations were performed according to D’Angelo et al. (2006). Fluorescence microscopy was performed with an inverted microscope (Leica DMIRE2) equipped with the Leica TCS SP2 laser-scanning device (Leica Microsystems GmbH, Wetzlar, Germany). Detection of fluorescence was performed for GFP with excitation at 488 nm (Ar/Kr Laser). Autofluorescence of plastids was detected at 650–720 nm. All images were acquired using a 63x/1.20 water-immersion objective (HCX PL Apo CS) from Leica.

Measurement of guard cell movements and ratiometric calcium analyses

Detached A. thaliana leaves from 3–5-wk-old soil-grown plants, grown under short-day conditions, were floated on incubation buffer (10 mM MES-KOH, pH 6.5, 10 mM KCl and 50 µM CaCl2) at 23–24°C. After 2 h of incubation the solution was replaced with incubation buffer with a final calcium concentration of 5 mM and the leaves were incubated for another 2 h. Subsequently, epidermal peels of each line were prepared, and the length and opening width of 50 stomata were determined using an inverted microscope. Each assay was performed in triplicate and as a blind experiment.

Cytosolic calcium transients in A. thaliana guard cells were analyzed in epidermal strips prepared from 3- to 6-wk-old soil-grown plants expressing the calcium indicator YC3.6. Epidermal strips were prepared by gently pressing the abaxial side of the leaf on a coverslip covered with a thin layer of medical adhesive (Hollister, Libertyville, USA) and removing the upper layers of the leaf with a razorblade. Samples of the different lines were then incubated for 2–4 h in incubation buffer (10 mM MES/Tris, 10 mM KCl and 50 µM Ca2+, pH 6.15) at 22°C in the light. Coverslips were then placed in a perfusion chamber which was fit to an inverted Leica DMI 6000B microscope equipped with an emission filter wheel (LUDL Electronic Products, Hawthorne, NY, USA) and an Orca ER cooled digital camera (Hamamatsu, Herrsching, Germany). Excitation was provided by a xenon lamp through a 440-nm filter. Emission filters were 485 nm (CFP) and 535 nm (YFP). Ratiometric images were collected and ratios of medium intensities in regions of interest (ROIs) calculated using openlab 5.2 software (Improvision, Tübingen, Germany). Pictures were taken at 6-s intervals. Cells showing weak fluorescence were excluded from data collection. Ratiometric measurements were performed when the guard cells of one stomate had opened nicely during the incubation time and showed equal size and form and full turgor for the measurement. Cells were monitored in the incubation buffer (continuous flow-through) for 5–15 min, and the buffer was then quickly replaced by incubation buffer that had a final calcium concentration of 5 mM to elicit endogenous calcium oscillations.

Stomatal aperture measurements after imposed calcium oscillations

Epidermal strips were prepared as described for ratiometric measurements. Intracellular calcium transients were imposed in cas-1 and wild-type guard cells by incubation of epidermal strips in depolarizing buffer (10 mM MES/Tris, pH 5.6, and 50 mM KCl) for 2 h and subsequent alternating 5-min pulses of hyperpolarizing buffer (10 mM MES/Tris, pH 5.6, 1 mM CaCl2 and 1 mM KCl) and depolarizing buffer. Stomatal apertures at time-point zero were considered to represent 100% for each experiment (n = 50 guard cells).


CAS – a conserved plant-specific calcium sensor protein

We recently identified the calcium sensor protein CAS in thylakoid membranes of C. reinhardtii after fractionation of isolated chloroplasts (Allmer et al., 2006). Comparative proteomics using tandem mass spectrometric peptide profiling validated the CAS plastid localization in C. reinhardtii (Naumann et al., 2007). The orthologous CAS protein from A. thaliana has, however, previously been reported to represent a cell surface receptor that mediates extracellular Ca2+ sensing in guard cells (Han et al., 2003; Tang et al., 2007). Conversely, but in line with the C. reinhardtii proteomic study (Allmer et al., 2006), an A. thaliana proteomic study also identified CAS as a chloroplast-localized thylakoid-associated protein (Friso et al., 2004). This situation prompted us to pursue a detailed investigation of the evolutionary conservation, cellular localization and function of the A. thaliana CAS protein.

To investigate the phylogenetic distribution of CAS we performed a database search using the CAS protein sequence from A. thaliana as query. This analysis identified 10 CAS-related protein sequences from nine different plant species (Fig. 1). No closely related protein sequence was detected in nonplant species. Conserved CAS proteins were detected in two species of the green alga genus Osteroccus, the moss Physcomitrella patens, and the fern Pteris vittata as well as in several monocotyledonous and dicotylodonous species. With the exception of P. patens, which possesses two closely related CAS proteins, this protein appears to be encoded by a single-copy gene in all other species.

Figure 1.

Phylogenetic analysis of Ca2+-sensing receptor (CAS) proteins. An unrooted phylogenetic tree of the 11 identified CAS proteins together with a rhodanese-like protein from Arabidopsis thaliana is shown. The optimal tree with the sum of branch length = 6.2 is presented. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) is indicated next to the branches. The tree is drawn to scale, with branch lengths illustrated using the same units as those of the evolutionary distances used to infer the phylogenetic tree. The evolutionary distances were computed using the Poisson correction method and are in units of the number of amino acid substitutions per site. All positions containing gaps and missing data were eliminated from the data set. There were a total of 251 positions in the final data set. Accession numbers of proteins are: AtCAS, Arabidopsis thaliana (At5g23060); MtCAS, Medicago trucatula (AC137819_22); VvCAS, Vitis vinifera (CAO67978); OsCAS, Oriza sativa (Os02g0729400); PsCAS, Picea sitchensis (ABK25223); PvCAS, Pteris vittata (ABD64881); PpCAS1, Physcomitrella patens (XP_001779309); PpCAS2, Physcomitrella patens (XP_001771640); CrCAS, Chlamydomonas reinhardtii (XP_001702364); OlCAS, Ostreococcus lucimarinus (XP_001418544); OtCAS, Ostreococcus tauri (CAL54944); out-group, A. thaliana rhodanese-like protein (AT4G01050).

CAS proteins differ in their lengths, ranging from 351 to 435 aa, with the Osteroccus tauri protein exhibiting an additional C-terminal extension of approx. 100 aa. The two predicted proteins from P. patens do not contain an N-terminal methionine and appear to lack the N-terminal domain present in all other CAS proteins. This difference may be caused by incorrect annotation of the genome sequence. A detailed sequence comparison of the identified CAS proteins revealed a rather high degree of conservation between aa positions 100 and 420 of the protein alignment (42–74% aa identity when comparing the vascular plants; Supplementary Material Fig. S1). As a common feature, all proteins harbor a conserved rhodanese-like domain in the C-terminal region (Supplementary Material Fig. S1). Although a function has not been assigned to this domain experimentally, it may have a regulatory rather than an enzymatic function, as the potential catalytic cysteine residue in all sequences is exchanged for an asparate (Bordo & Bork, 2002) (Supplementary Material Fig. S1).

To further elucidate the potential subcellular localization of A. thaliana CAS protein and the newly identified CAS proteins from fern and seed plant species, we analyzed the protein sequences using established localization prediction algorithms (predotar, chlorop, and psort). In these analyses all CAS proteins were unambiguously classified as chloroplast-targeted proteins and characteristic N-terminal chloroplast-targeting sequences were identified. These target sequences represent the variable N-terminal domain of the CAS proteins and vary in length from 16 aa in P. vittata to 71 aa in Vitis vinifera. Moreover, the CAS protein sequences from all vascular plant species analyzed start in the N-terminal region with the canonical aa motive consisting of the amino acid combination methionine and alanine (MA) at the very N-terminus of the sequence (Bruce, 2000). Taken together, these results identify CAS as a conserved plant-specific protein that exhibits structural features typical for chloroplast-targeted protein localization.

Arabidopsis thaliana CAS is localized to chloroplast thylakoids

We next addressed the expression pattern of CAS by quantitative real-time PCR using RNA isolated from root, leaf, stem, flower and silique tissue, respectively. While no significant amounts of CAS mRNA were detected in root tissues, we observed strong CAS expression in leaves (Fig. 2). In stem, flower and silique tissues, CAS gene expression was also clearly detectable although at lower levels than in leaves. This expression pattern coincides with the distribution of photosynthetically active chloroplasts in plant tissues and is in accordance with the results of publicly available transcriptomics analyses presented at the eFP Browser (http://www.bar.utoronto.ca/efp/cgi-bin/efpWeb.cgi; Winter et al., 2007).

Figure 2.

Analysis of the Ca2+-sensing receptor (CAS) expression pattern. A tissue-specific quantitative real-time polymerase chain reaction analysis of CAS mRNA levels was performed. Specific primers for the CAS coding region were used to amplify cDNAs from RNA isolated from Arabidopsis thaliana plant tissues. The relative expression of the CAS gene was calculated using Actin 2 (ACT2) as internal reference.

To experimentally investigate the subcellular localization of CAS from A. thaliana, we determined the subcellular localization of CAS::GFP fusion proteins in transiently transformed N. benthamiana protoplasts using confocal laser microscopy (Fig. 3a). As localization controls we included constructs expressing GFP, TPT::GFP (triose-phosphate/phosphate translocator; Schneider et al., 2002) and CBL1::GFP (D’Angelo et al., 2006) as established cytoplasmic, chloroplast and plasma membrane localization markers, respectively. We injected A. tumefaciens cultures expressing the different constructs into intact plant leaves and isolated protoplasts 3 d after inoculation. Localization of the GFP::CAS fusion protein was exclusively observed in chloroplasts (Fig. 3a). An identical localization was detected for the GFP::TPT fusion protein. Importantly, deletion of the N-terminal transit peptide (CASΔN) abolished chloroplast targeting of the fusion protein and resulted in cytoplasmic localization, as was observed for the GFP protein. Investigation of CBL1::GFP illustrated the localization of a plasma membrane-localized protein.

Figure 3.

Analysis of the subcellular localization of the Ca2+-sensing receptor (CAS). (a) Subcellular localization of green fluorescent protein (GFP) fusion proteins was analyzed in protoplasts isolated from transiently transformed tobacco (Nicotiana benthamiana) leaves. The identity of the different fusion proteins is indicated on the right. The left-hand panels show GFP fluorescence microscopy images, the central panels show chloroplast autofluorescence images and the right-hand panels present overlays of the two images. Triose-phosphate/phosphate translocator (TPT) and Calcinevrin-B like protein 1 (CBL1) were used as chloroplast and plasma membrane markers, respectively. (b) Localization studies of CAS::GFP fusion proteins in epidermal cells of N. benthamiana leaves using laser-scanning microscopy. The left-hand column shows GFP fluorescence images, and chloroplast autofluorescence and overlay images are displayed in the two middle columns. The right-hand column presents a bright field image of the analyzed epidermal section. CAS::GFP and TPT:GFP fusions are targeted to the stomatal chloroplasts.

As previous studies have reported that CAS is involved in stomatal regulation (Han et al., 2003), we were especially interested in investigating the localization of CAS::GFP fusion proteins in guard cells of intact leaves. To this end, transiently transformed leaf discs from N. benthamiana were examined by confocal laser microscopy. In these experiments we detected CAS::GFP as well as the TPT::GFP fusion proteins exclusively in the chloroplasts of the two guard cells (Fig. 3b). No fluorescence was observed in the cytosol or at the plasma membrane (Fig. 3b). In addition, we analyzed the localization of CAS::GFP in photosynthetic parenchyma tissue and solely detected this fusion protein in chloroplasts, and not in other cellular compartments (data not shown). These experiments provide evidence for a function of CAS in chloroplasts and do not support the localization of this calcium sensor protein in other cellular compartments.

To further corroborate these cell biological analyses of CAS localization we performed subcellular fractionation analyses by isolating intact chloroplasts from transiently transformed tobacco plants expressing a CAS::StrepII tagged fusion protein. Isolated chloroplasts were further separated into thylakoid and soluble fractions. Western blot analysis with a StrepII-specific antibody revealed a strong enrichment of the CAS protein in the thylakoid subfraction (Fig. 4). Immunoblot detection of the thylakoid marker protein chlorophyll binding protein D1 (PsbA) validated the thylakoid localization of CAS (Fig. 4).

Figure 4.

The Ca2+-sensing receptor (CAS) is a thylakoid-located protein. Western blot analysis of isolated chloroplasts from tobacco (Nicotiana benthamiana) leaves transiently transformed with CAS::StrepII fusion constructs was performed. Whole chloroplasts (total) and thylakoid membranes were isolated and separated on 12% SDS gels and analyzed by western blot. The CAS protein was detected using StrepII-specific antibodies, while the control protein chlorophyll binding protein D1 (PsbA) was detected with protein-specific antibodies. For each fraction, 15 µg of protein was used.

Loss of CAS function impairs Ca2+-induced stomatal closure

To elucidate the function of CAS in plants, we employed a reverse-genetics approach and isolated and characterized two T-DNA insertion mutant lines. The mutant alleles cas-1 (SALK_070416) and cas-2 (GABI_665G12) were obtained as segregating lines from the NASC and GABI-KAT collections, respectively. Homozygous individuals for both alleles were identified by PCR using T-DNA- and gene-specific primer combinations. Sequence analysis of PCR products generated with primer combinations specific for the flanking regions of the T-DNA insertions localized the T-DNA in cas-1 230 bp downstream of the ATG initiator codon within the first intron of the gene (Fig. 5a). In cas-2 the T-DNA insertion was located within the third intron and this insertion event turned out to be accompanied by a deletion of a 38-bp genomic sequence. Therefore, one border of the T-DNA follows nucleotide 819 of the CAS genomic sequence, whereas the other border precedes nucleotide 857 (Fig. 5a).

Figure 5.

T-DNA insertions in the Ca2+-sensing receptor (CAS) gene generate loss-of-function alleles. (a) Intron–exon structure of the Arabidopsis thaliana CAS gene (coding region) and location of the T-DNA insertion. Gray boxes and connecting lines depict exons and introns, respectively. The insertion sides of both T-DNAs are indicated schematically. The T-DNA in cas-1 (SALK_070416) is located in the first intron, while in cas-2 (GABI_665G12) the insertion is located at the intersection of exon 3 and intron 3. Δ38 illustrates the deletion of 38 nt. The location and orientation of primers used in quantitative real-time polymerase chain reaction (PCR) experiments are shown as black arrows. (b) CAS and Actin 2 (ACT2) transcript levels as determined by real-time PCR in the wild-type (WT) and cas-1 and cas-2 mutant lines. CAS transcript levels were analyzed for both full-length mRNA regions and for the transcribed region downstream of both T-DNA insertions (denoted as CAS and CAS3′, respectively). Both T-DNA insertions reduce CAS mRNA to nondetectable levels.

To determine CAS mRNA expression in the two T-DNA insertion lines we performed gene-specific RT-PCRs on homozygous mutant individuals and wild-type control plants. PCR transcript analysis of the constitutively expressed gene Actin 2 (ACT2) served as an internal control. A CAS-specific transcript was clearly detected in wild-type plants (Fig. 5b). By contrast, expression of functional CAS mRNA was reduced to undetectable levels by the T-DNA insertion in both cas-1 and cas-2 mutant lines. Moreover, RT-PCR with a primer combination specific for the region downstream of the insertions led to the generation of a PCR product in wild-type control plants, but not in cas mutant lines (Fig. 5b). These results indicate that CAS expression in the two mutant alleles is reduced to undetectable levels. The isolated homozygous individuals as well as their progeny showed no obvious phenotype under normal growth conditions.

Previous studies in A. thaliana guard cells established the importance of calcium signaling for the regulation of stomatal aperture. Elevation of external calcium concentration promotes stomatal closure (Allen et al., 1999, 2000, 2001). Previous investigations of this response reaction using CAS antisense lines suggested a crucial role of CAS in mediating the closure in response to external calcium application (Han et al., 2003). Therefore, we were especially interested to comparatively investigate this stomatal response reaction in cas-1 and cas-2 mutant and wild-type plants. Application of 5 mM CaCl2 to intact leaves significantly reduced the stomatal aperture in wild-type plants within 2 h (Fig. 6). However, in the cas-1 and cas-2 mutants, stomatal closure was significantly attenuated compared with the wild-type response (Fig. 6). In control treatments without CaCl2 application we did not observe differences in stomatal aperture when comparing wild-type and mutant plants (Fig. 6). Comparative analyses of stomatal closure in response to abscisic acid (ABA) application did not reveal differences between the cas mutants and wild type (data not shown). These results indicate that the chloroplast-localized protein CAS is specifically involved in regulating stomatal closure in response to external calcium elevation.

Figure 6.

Loss of Ca2+-sensing receptor (CAS) function impairs stomatal responses to elevation of the exogenous Ca2+ concentration. The Ca2+-induced stomatal closure responses of Arabidopsis thaliana wild-type (WT; open bars) and cas-1 (a) and cas-2 (b) (closed bars) mutants were analysed. Leaves from wild-type and cas mutant plants were exposed to the indicated calcium concentrations for 2 h. After preparation of epidermal peels, the stomatal aperture was analyzed by determining the ratio of the length and opening width of the stomatal pore. The ratio was calculated from the data of three independent experiments, each encompassing values from 50 stomata. Error bars indicate the SE.

CAS function is crucial for the generation of cytoplasmic Ca2+ signals

Stomatal closure in response to increases in the external Ca2+ concentration is accompanied by transient and repetitive elevations in [Ca2+]cyt (Allen et al., 2000). To investigate the contribution of the chloroplast-localized CAS protein to Ca2+-dependent regulation of stomatal closure in more detail, we further analyzed cytoplasmic calcium dynamics in response to exogenously applied calcium. To this end we generated wild-type as well as cas-1 lines expressing the fluorescent calcium indicator YC3.6 (Nagai et al., 2004) under the control of the 35S promoter. Transgenic cas-1 and wild-type plants expressing YC3.6 were subjected to ratiometric fluorescence calcium measurements after Ca2+ exposure. Application of 5 mM calcium to epidermal peels of A. thaliana leaves induced a strong calcium spike within 30 s in wild-type guard cells (Fig. 7a). This initial calcium elevation was followed by consecutive Ca2+ oscillations that proceeded for the duration of the measurements (15 to 20 min). With increasing time after calcium application we observed a decline in the amplitudes of oscillations of [Ca2+]cyt. In sharp contrast, the cas-1 mutant exhibited only very weak calcium oscillations and in most of our analyses (> 80% of measurements) we did not observe an initial Ca2+ spike (Fig 7b). These results are largely congruent with the previous finding that cytoplasmic oscillations in Ca2+ concentration were not detectable in CAS antisense plants after exposure to external Ca2+ stimuli (Han et al., 2003). Moreover, these data provide evidence for a crucial role of the chloroplast-localized CAS protein in the generation and/or regulation of cytoplasmic Ca2+ responses.

Figure 7.

The Ca2+-sensing receptor (CAS) regulates calcium oscillations in guard cells. Altered oscillation kinetics of intracellular calcium transients in the cas-1 mutant were found. (a) Induced cytoplasmic Ca2+ concentration ([Ca2+]cyt) oscillations in a wild-type (WT) guard cell elicited by an external Ca2+ concentration of 5 mM (black bar). Insets depict ratio images of the corresponding guard cell complex at time-points before, during and after a transient. (b) Induced [Ca2+]cyt oscillations in a cas-1 guard cell elicited by an external Ca2+ concentration of 5 mM (black bar). Insets depict ratio images of the corresponding guard cell complex before and at the moment of the (compared to the WT) initial transient. The response of the internal calcium oscillations to external elevations in calcium concentrations is strongly impaired in cas mutant lines.

We next addressed the questions of whether the signaling events downstream of the intracellular Ca2+ oscillations are also affected in cas mutants and whether the impairment of cytosolic Ca2+ elevations is causally linked to the observed differences in stomatal responsiveness. We applied externally imposed Ca2+ oscillations to epidermal strips and measured stomatal closure at several time-points. This method was established by Allen et al., (2001) and allows one to determine whether a mutant is still able to respond to the correct Ca2+ signal or whether its defect lies downstream of the cytosolic Ca2+ elevations. Four 5-min pulses of 1 mM Ca2+ buffer under hyperpolarizing conditions were applied to epidermal strips incubated in Ca2+-free depolarizing buffer, thereby imposing four externally controlled Ca2+ transients on the guard cells (Allen et al., 2000, 2001). The wild-type and cas plants responded to these imposed cytosolic Ca2+ oscillations with similar degrees of stomatal closure (Fig. 8). Stomatal aperture was reduced to approx. 60% of its initial value after 70 min, while the time-course was comparable between wild-type (Fig 8a) and cas-1 mutant plants (Fig 8b). This result clearly indicates that the mutant is able to respond to Ca2+ signals and provides further evidence that the CAS protein is necessary to generate the Ca2+ transients required for stomatal closure.

Figure 8.

Imposed calcium oscillations induce stomatal closure in Ca2+-sensing receptor 1 (cas-1) mutants. Intracellular calcium transients were imposed in cas-1 and wild-type (WT) guard cells by incubation of epidermal strips in depolarizing buffer (white bars) and subsequent alternating 5-min pulses of hyperpolarizing (black bars) and depolarizing buffer. Both the cas-1 mutant guard cells (a) and the WT guard cells (b) responded to the imposed transients by stomatal closure to similar extents. Initial stomatal apertures of 100% correspond to 2.95 ± 0.11 µm and 4.32 ± 0.14 µm for cas-1 and WT, respectively. Error bars depict mean ± SEM (n = 50 guard cells).


The CAS protein is localized in chloroplast thylakoid membranes

The scope of this study was to elucidate the subcellular localization of the CAS protein and to investigate its physiological functions and potential role in guard cell regulation and signaling. Our study identifies CAS as a plant-specific protein that is encoded by a single-copy gene and is conserved in algae, mosses, ferns and seed plants. Importantly, the identified predicted full-length proteins from all analyzed species contain an N-terminal domain that is clearly predicted to function as a chloroplast transit peptide. Our expression analysis of CAS from A. thaliana revealed a predominant expression of this gene in tissues that harbor photosynthetically active chloroplasts. The observed level of CAS gene expression coincided with the amount of chloroplasts per cell in the tissues investigated, with the highest amounts of mRNA detected in leaves and moderate expression in stems, flowers and siliques. We investigated the localization of GFP fusion proteins to address the cellular localization of CAS experimentally. These studies revealed an exclusive chloroplast localization of CAS in mesophyll cells as well as in guard cells. In none of our experiments did we detect traces of CAS::GFP fluorescence outside of the chloroplasts. Deletion of the N-terminal transit peptide abolished chloroplast localization of the analyzed GFP fusion protein and thereby supported the functional significance of the N-terminus for proper CAS localization. We further corroborated our cell biological analyses of CAS localization with biochemical approaches that involved subcellular fractionation analyses. These experiments further substantiated the chloroplast localization of CAS and identified this protein as being enriched in thylakoid membranes. This facet of our results is difficult to reconcile with a recently published report that suggested a plasma membrane localization of CAS to enable a function as an extracellular calcium sensor protein (Han et al., 2003). In this work, CAS localization was analyzed by fluorescence microscopic inspection of CAS::GFP fusion proteins that were transiently expressed in onion (Allium cepa) epidermal cells. As these cells do not possess differentiated chloroplasts, their investigation without confocal microscopy may have blurred potential fluorescence originating from small amyloplast organelles. However, our results are in line with several reported proteomic studies that identified CAS-derived peptides in purified thylakoid fractions from A. thaliana and C. reinhardtii (Friso et al., 2004; Peltier et al., 2004; Allmer et al., 2006; Naumann et al., 2007). Moreover, while this work was under consideration, an independent study by Nomura et al. also reported a chloroplast localization of the CAS protein and similar phenotypes of CAS loss-of-function alleles (Nomura et al., 2008). In summary, we conclude that CAS represents a conserved protein that primarily functions in thylakoid membranes of chloroplasts in higher plants and algae.

A chloroplast protein is crucial for the regulation of stomatal movements and the generation of cytoplasmic Ca2+ transients

With the exception of the orchid genus Paphiopedilum, guard cells of all plant species studied so far have been shown to contain chloroplasts (D’Amelio & Zeiger, 1998). Guard cell chloroplasts possess functional photosystems I and II and electron transport, oxygen evolution and photophosphorylation have been clearly demonstrated (Zeiger et al., 2002). The interconnection of chloroplast physiology and guard cell responsiveness to environmental cues has been a focus of controversy (Zeiger et al., 2002; Baroli et al., 2008). However, it has been firmly established that guard cell signaling and movements in response to, for example, blue light, CO2 and ABA can function independently of photosynthetic activity (Roelfsema et al., 2002; von Caemmerer et al., 2004).

In this study, we report that loss of CAS function impairs guard cell closure that is induced by elevation of the extracellular Ca2+ concentration. The finding that two independent cas mutant alleles confer similar failures of responsiveness clearly links this phenotype to CAS function. Responsiveness to extracellular Ca2+ elevation was not completely diminished in the cas loss-of-function mutants, suggesting that the chloroplast-localized CAS protein contributes to an important but not necessarily essential facet of the elaborate signaling network regulating stomatal movements. We did not observe differences in stomatal movements between wild-type and mutant lines in response to ABA, indicating that CAS function in mediating stomatal closure is rather specific and does not reflect pleiotropic effects of impairments in chloroplast physiological function. In this regard it is also noteworthy that we did not observe obvious phenotypic differences between mutant and wild-type plants in regular growth conditions and that cas mutants and wild-type plants did not exhibit differences in chlorophyll content and fluorescence (data not shown). However, as the experimental elevation of extracellular Ca2+ represents a rather artificial stimulus for modulating stomatal aperture it will be very interesting to investigate the regulation of guard cells in cas mutants in response to physiological cues such as humidity, CO2, O3 and light/dark transitions in the future.

Our investigation of cytoplasmic Ca2+ dynamics in response to extracellular Ca2+ exposure using plant lines that constitutively expressed the reporter protein YC3.6 clearly revealed a crucial role of CAS in intracellular Ca2+ mobilization and regulation. While an increase of extracellular Ca2+ to 5 mM induced rapid and repetitive Ca2+ spiking in guard cells of wild-type plants, this response was strongly diminished in mutant plants. This finding suggests a critical function of chloroplasts in generating cytoplasmic calcium transients. Imposition of Ca2+ oscillations by membrane hyperpolarization was sufficient to induce stomatal closure in wild-type and cas mutant plants. This result indicates the impairment of cytoplasmic Ca2+ oscillations as the cause of the observed stomatal phenotype resulting from loss of CAS function. However, our data do not allow mechanistic insights into how CAS function contributes to the modulation of [Ca2+]cyt. A direct association of this thylakoid membrane protein with channels or transporters that could mediate Ca2+ release from the chloroplast appears rather unlikely. Nevertheless, it is tempting to speculate that the CAS protein could function in Ca2+ storage and accumulation in chloroplasts and thereby would enable a rapid release of Ca2+ into the cytoplasm. However, while the contribution of the endogenous calcium stores vacuole and endoplasmatic reticulum has been intensively studied, the potential contribution of chloroplasts to cellular Ca2+ dynamics has remained largely underappreciated (Johnson et al., 2006). Our identification of the chloroplast-localized CAS as a critical component of guard cell regulation and cytoplasmic Ca2+ release opens a novel research direction in cellular Ca2+ signaling. Nonetheless, further analyses of CAS function will need to consider that this protein is not only present in guard cells but is also likely to perform critical functions in mesophyll and epidermal cells.

Chloroplasts function in cellular calcium signaling

It has long been established that chloroplast-localized physiological processes are subject to regulation by Ca2+. Regulation of the enzyme activity, assembly and stability of the photosystem II/oxygen-evolving complex are all regulated by Ca2+ (Johnson et al., 2006). Although chloroplasts have been shown to contain high concentrations (4–43 mM) of Ca2+ and although there has been ample evidence for fluctuation in chloroplast Ca2+ concentration, the interconnection of the Ca2+ dynamics of this organelle with the cellular Ca2+ signaling network has not been addressed. The surprising identification of the two chloroplast-localized channel proteins CASTOR and POLLUX as essential components for cytoplasmic Ca2+ oscillation during root nodulation has provided the first challenge to the theory of an isolated regulation of chloroplast calcium dynamics (Imaizumi-Anraku et al., 2005). Our identification of CAS as a crucial protein for cytoplasmic Ca2+ responses in guard cells will enable the further exploration of the contribution of chloroplasts to Ca2+ signaling.


We thank Dr Ulf-Ingo Flügge for providing the TPT::GFP fusion construct and Professor Julian Schroeder for providing the YC3.6 expression plasmid. We are indebted to Professors Alistair Hetherington and Julian Schroeder for advice regarding the guard cell bioassays. This work was supported by grants from the DFG (JK and MH) and DAAD (JK).