A toolset of aequorin expression vectors for in planta studies of subcellular calcium concentrations in Arabidopsis thaliana

Abstract

Calcium has long been acknowledged as one of the most important signalling components in plants. Many abiotic and biotic stimuli are transduced into a cellular response by temporal and spatial changes in cellular calcium concentration and the calcium-sensitive protein aequorin has been exploited as a genetically encoded calcium indicator for the measurement of calcium in planta. The objective of this work was to generate a compatible set of aequorin expression plasmids for the generation of transgenic plant lines to measure changes in calcium levels in different cellular subcompartments. Aequorin was fused to different targeting peptides or organellar proteins as a means to localize it to the cytosol, the nucleus, the plasma membrane, and the mitochondria. Furthermore, constructs were designed to localize aequorin in the stroma as well as the inner and outer surface of the chloroplast envelope membranes. The modular set-up of the plasmids also allows the easy replacement of targeting sequences to include other compartments. An additional YFP-fusion was included to verify the correct subcellular localization of all constructs by laser scanning confocal microscopy. For each construct, pBin19-based binary expression vectors driven by the 35S or UBI10 promoter were made for Agrobacterium-mediated transformation. Stable Arabidopsis lines were generated and initial tests of several lines confirmed their feasibility to measure calcium signals in vivo.

Aequorin, calcium measurements, calcium signalling, subcellular localization, YFP-fusion protein

Introduction

Calcium is an important signalling molecule involved in the regulation of a vast number of cellular processes (reviewed in DeFalco et al., 2010; Dodd et al., 2010; Kudla et al., 2010). In response to a variety of abiotic and biotic stimuli, calcium is released from extracellular and intracellular calcium stores, thereby forwarding the environmental information into a temporary and spatially controlled reaction to the stimulus. It was hypothesized that the specificity of the calcium signal is achieved by its shape, intensity, and length, as well as by repetition of the calcium release (Dolmetsch et al., 1998; Li et al., 1998; McAinsh and Pittman, 2009). In order for the specific signal to be created, calcium must be released from storage compartments in an orchestrated manner, unique for each stimulus. Subsequently, receptor molecules sense the change in calcium concentration and mediate the signal into an appropriate cellular response. This can either happen directly or the receptor molecule can be the starting point of a cellular signalling cascade. Calcium released from storage compartments initially results in a high calcium concentration next to the membrane from where it is emitted (Etter et al., 1996). Membrane-associated calcium receptor proteins such as myristoylated and palmitoylated CDPKs (calcium-dependent protein kinases) can receive the signal directly at the membrane and forward the signal by protein phosphorylation (Lu and Hrabak 2002, Dammann et al., 2003; Benetka et al., 2008; Mehlmer et al., 2010). Other calcium sensors are soluble proteins and can sense the signal anywhere within the compartment in which the calcium is released (reviewed in Kudla et al., 2010).

Calcium regulation is best described for the cytoplasm where it is known to regulate diverse cellular processes such as metabolic pathways, ion channels or gene transcription (reviewed in Reddy et al., 2011). There is growing evidence, that calcium also affects organellar processes such as chloroplast functions (Jarrett et al., 1982; Roberts et al., 1983; Chigri et al., 2005, 2006; Bussemer et al., 2009a, b; Bayer et al., 2011; Stael et al., 2011). It is also known that calcium is released into the chloroplast stroma after the transition from light to dark (Johnson et al., 1995). Moreover, it is known that illuminated chloroplasts have the ability to incorporate large amounts of calcium (Nobel, 1967; Muto et al., 1982). Nevertheless, the understanding of calcium regulation within organelles is still scarce. A comprehensive overview about the current knowledge on plant organellar calcium signalling was recently provided by Stael et al. (2012).

Calcium can be detected by fluorescent dyes but the main difficulty is to get them into the plant cell by passing the cell wall and the plasma membrane (Lee et al., 1999). In addition, calcium-detecting dyes are widely distributed in the cell and cannot easily be used to discriminate between different compartments. Specific subcellular localization can be achieved by utilization of the protein-sorting machinery provided that the calcium detector is a protein. Two different genetically encoded calcium indicators (GECIs) are mainly used as calcium sensors within biological systems. In fluorescence-based GECIs, calcium binding induces a conformational change that results in the approximation of the fluorophores creating a detectable FRET signal (Miyawaki et al., 1997; Mank and Griesbeck 2008; Tian et al., 2009; Knöpfel et al., 2010). Unfortunately, the readout requires high light intensities which are not suitable for all applications, especially when analysing chloroplasts. By contrast, aequorin is a luminescent GECI, emitting light directly upon calcium binding (Shimomura et al., 1962; Plieth, 2006; Shimomura, 2006). Calcium-dependent light emission from aequorin can be used to detect changes in calcium concentration in the compartment where aequorin is present. Aequorin has been exploited for the measurements of subcellular calcium changes by fusion to targeting sequences for different sub-subcellular compartments (Knight et al., 1991; Johnson et al., 1995; Mithofer and Mazars, 2002). Nevertheless, the binary plasmids used to create these plant lines were designed by different groups for unique experiments using different binary vector backbones for T-DNA integration.

In this work, a compatible set of expression vectors is described that target aequorin to different subcellular localizations using an identical vector backbone. Expression can be driven by either the 35S or the UBI10 promoter and two different resistant cassettes (kanamycin and Basta) are available for selection. All constructs consist of a YFP-aequorin (YA) fusion that allows easy verification of the correct subcellular localization. It can also be used to confirm expression of the constructs and can assist in the selection of transgenic plant lines by fluorescent microscopy. The set includes constructs targeted to the cytosol, nucleus, the inner surface of the plasma membrane, mitochondria, and the stroma, as well as the inner and outer surface of the chloroplast envelope membranes. Modular set-up of the vectors allows the exchange of targeting sequences and thus the expression system can easily be customized for other subcellular compartments. Analysis of several stable Arabidopsis lines confirmed their feasibility to measure calcium signals in vivo.

Material and methods

Molecular cloning and construction of expression plasmids

YA fusion constructs were placed into the pRT100 vector under control of the 35S promoter (Topfer et al., 1987; Benetka et al., 2008). For that purpose, the coding region of a GFP-aequorin fusion construct (a kind gift from Chris Bowler; Ecole Normale Superieure, Paris, France) was amplified by PCR with a forward primer to introduce an ApaI and a NotI restriction site and a reverse primer to introduce an XbaI site (see Supplementary Table S1 at JXB online). The PCR product was subsequently cloned via the restriction sites ApaI and XbaI into a modified pRT100. The coding region of GFP was subsequently replaced by eYFP (Clontech 1290 Terra Bella Ave. Mountain View, CA 94043 USA) and the new plasmid was termed p35S-YA (see Supplementary Material S1 at JXB online). To create the plasmid pUBI-YA, the 35S promoter was replaced via the restriction sites KpnI/ApaI by the ubiquitin promoter (UBI10) amplified by PCR from Arabidopsis thaliana genomic DNA (see Supplementary Material S1 and Supplementary Data at JXB online).

Most targeting sequences were amplified by PCR using primers to introduce an ApaI and a NotI restriction site (see Supplementary Table S1 at JXB online) and subsequently introduced into both p35S-YA and pUBI-YA. In the case of the NLS- and NES-YA fusion proteins, targeting sequences were included into the forward primer (after the NotI restriction site) and YA was amplified together with the YA-reverse primer (see Supplementary Table S1 at JXB online). Both PCR-products, NLS-YA and NES-YA, were then used to replace YA via ApaI/XbaI restriction. In the case of NES, the coding sequence of a cytosolic protein was additionally fused in front of the NES-YA by ApaI/NotI restriction. More detailed information on the chosen targeting sequences can be found in the supplementary data (see Supplementary Material S1 at JXB online).

To create plasmids suitable for Agrobacterium-mediated transformation of plants, the whole expression cassettes of the different constructs (Fig. 1) were cloned with the restriction enzymes KpnI and SacI for p35S-YA or KpnI and SpeI for pUBI-YA into pBIN19 (kanamycin) and pBIN19 (Basta/Bar) (Bevan, 1984). An overview over all final constructs is presented in Table 1. All vector sequences can be found in the supplementary data (see Supplementary Material S2 at JXB online).

Table 1.

Overview over the different binary YA fusion vectors for stable transformation

Subcellular localization	Binary plasmid 35S promoter	UBI10 promoter
Cytosol and nucleus	pBIN-YA(K)	pBINU-YA(K)
	pBIN-YA(B)	pBINU-YA(B)
Cytosol only	pBIN-CYA(K)	pBINU-CYA(K)
	pBIN-CYA(B)	pBINU-CYA(B)
Nucleolus only	pBIN-NYA(K)	pBINU-NYA(K)
	pBIN-NYA(B)	pBINU-NYA(B)
Plasma membrane	pBIN-PMYA(K)	pBINU-PMYA(K)
	pBIN-PMYA(B)	pBINU-PMYA(B)
Mitochondria	pBIN-MYA(K)	pBINU-MYA(K)
	pBIN-MYA(B)	pBINU-MYA(B)
Chloroplast OE surface	pBIN-OEYA(K)	pBINU-OEYA(K)
	pBIN-OEYA(B)	pBINU-OEYA(B)
Chloroplast IE surface	pBIN-IEYA(K)	pBINU-IEYA(K)
	pBIN-IEYA(B)	pBINU-IEYA(B)
Stroma	pBIN-CHYA(K)	pBINU-CHYA(K)
	pBIN-CHYA(B)	pBINU-CHYA(B)

For each variant, a 35S promoter-based (pBIN) and a UBI10 promoter-based (pBINU) version was constructed. Both versions are available as vectors carrying either kanamycin (K) or Basta (B) resistance.

Fig. 1.

Overview of the cloning strategy for the creation of binary expression vectors targeting YA to different cellular compartments. All targeting sequences were cloned into an expression cassette in front of YA using the restriction enzymes ApaI and NotI. The complete expression cassette was subsequently transferred into the binary vector pBIN19.

Transient expression of fusion constructs in Brassica rapa and Nicotiana benthamiana

Brassica rapa protoplasts were isolated and transiently transformed with all fusion constructs in p35S-YA as described by Mehlmer et al. (2011). Agrobacterium-mediated transformation of tobacco leaf cells was performed with pBINU-MYA as described in Voinnet et al. (2003) with the Agrobacterium strain LBA1334. Protoplasts were isolated out of leaf tissue 48 h after transformation as described in Koop et al. (1996) and further analysed by confocal laser scanning microscopy.

Confocal laser scanning microscopy

In vivo localization of YA-fusion proteins was performed using a Leica TCS SP5 confocal laser scanning microscope. YFP emission was excited with the argon laser at 514 nm and the emitted fluorescence was detected between 525–546 nm. Chlorophyll fluorescence was also excited at 514 nm and measured between 657 and 726 nm. To detect mitochondria, protoplast suspensions were incubated for 30 min at room temperature with 125 nM of MitoTracker (RedCMX Ros Invitrogen, 1 mM stock in dimethyl sulphoxide) in the appropriate protoplast incubation media. Protoplasts were re-suspended occasionally with gentle agitation. After centrifugation at 500 g for 2 min without engaging the break, the supernatant was discarded and the protoplasts were carefully resuspended in protoplast incubation media. The washing step was repeated three times. MitoTracker RedCMX Ros emission was excited with the DPSS LASER at 561 nm and measured between 585–632 nm. Image processing was performed using the Leica Application Suite for Advanced Fluorescence (LAS AF).

Chloroplast isolation and thermolysin treatment

YA fusion proteins were transiently expressed in Nicotiana tabacum according to Voinnet et al. (2003). Two days after infiltration, the expression was confirmed by fluorescence microscopy and leaves were harvested. All following steps were performed at 4 °C. Using a small blender, leaves were homogenized in isolation buffer (330 mM sorbitol, 20 mM MOPS, 13 mM TRIS, 3 mM MgCl₂, 0.1% BSA, pH 7.9) and filtered through a nylon mesh (30 μm diameter). After centrifugation at 2000 g for 10 min the crude chloroplast fraction was resuspended in wash buffer (300 mM sorbitol, 50 mM HEPES-KOH, 3 mM MgCl₂, 1 mM CaCl₂, pH 7.6). Thermolysin was added to the chloroplast suspension to a final concentration of 0.1 mg ml⁻¹ and incubated for 20 min on ice. The reaction was stopped by the addition of 5 mM EDTA. Chloroplasts were recovered by centrifugation at 2000 g for 5 min and analysed by SDS-PAGE and Western blotting using antisera against α-aequorin (Abcam, UK) and α-GFP (Roche, Germany).

Generation of transgenic plants lines

Transgenic Arabidopsis plant lines were generated using the Agrobacterium strain GV3101 according to the procedure described in Clough and Bent (1998). Plants were selected by growth on MS-plates (Murashige and Skoog medium with 0.8% plant agar; Duchefa Biochemie, Haarlem, The Netherlands; Murashige and Skoog, 1962) containing 50 μg ml⁻¹ kanamycin and expression of the fusion constructs was verified by fluorescence microscopy.

Aequorin activity measurements

Seeds of transgenic plants were surface-sterilized by treatment with 70% ethanol for 1 min, placed on MS-plate (see above, Melford Laboratories Ltd., Ipswich, UK) without kanamycin and stratified at 4 °C for 4 d. Seedlings were grown for 7 d at 20 °C with a day/night cycle of 16/8 h. After 7 d, aequorin was reconstituted by treating the seedlings overnight in water with 10 μM coelenterazine (native, Cat. No. C-7001, Biosynth AG, Staad, Switzerland). Cold treatment was performed using a Peltier cooled metal plate according to Knight and Knight (2000). Calcium-induced light emission was recorded using an intensified CCD camera system (Photek 216, Photek Ltd., Hastings, UK) (Evans et al., 2005). Absolute calcium concentrations reported by each construct in planta were calculated using an empirically-determined calibration curve (Knight et al., 1996). This necessitates calculation of rate constant of aequorin activity at every time-point. To calculate rate constants, the theoretical maximum aequorin activity for each time-point was recorded by determining how much reconstituted aequorin was present. At the end of each experiment seedlings were frozen to -15°C to discharge this aequorin and subsequently warmed up again to 15 °C. Luminescence was measured continually to determine empirically how much aequorin activity was remaining (crucial for determination of maximal rates). Statistical significance of differences in peak ratios was assessed by performing a standard Student t test analysis assuming unequal variances (significance taken as P <0.05).

Results

Generation of YA fusion proteins targeted to different cellular subcompartments

The aim of this work was to create a set of compatible and comparable constructs that would allow the targeting of aequorin into different cellular subcompartments. To that end, the coding sequence of aequorin was fused to different signal peptides or protein domains that were selected to assist in the targeting process. Furthermore, to allow easy confirmation of subcellular localization, the constructs contained the coding sequence for YFP. Figure 1 shows a general scheme of the expression cassette utilized in the different constructs. Expression is driven from either the 35S or the UBI10 promoter. Targeting sequences are localized at the N-terminus of the fusion protein with the YFP placed between the targeting sequence and the aequorin. Putting the YFP in the middle instead of at the very end of the fusion protein resulted in a more faithful targeting of some of the constructs (see Supplementary Table S2 at JXB online) and was therefore selected as a general set-up. This also leaves the C-terminal proline of aequorin free, a property necessary for stable enzyme activity (Watkins and Campbell, 1993). Transcription in all constructs is terminated by the CaMV terminator.

Two constructs were generated to achieve cytoplasmic localization. The first construct solely contained the YA fusion without any targeting sequence. In a second construct, an exclusive cytosolic localization was attempted by introducing the nuclear export signal (NES) from the heat-stable protein kinase inhibitor (Wen et al., 1995) between the cytosolic CPK17_G2A (Benetka et al., 2008; Myers et al., 2009) and the YA. By contrast, the nuclear localization sequence (NLS) of the SV40 protein (Kalderon et al., 1984) was placed before YA to achieve exclusion from the cytosol. The first 58 amino acids of the calcium-dependent protein kinase 17 (CPK17) were selected to direct YA to the plasma membrane based on the strong N-myristoylation and palmitoylation of this peptide (Benetka et al., 2008). The first 66 amino acids of the mitochondria-localized 2-oxoglutarate dehydrogenase E1 (AKDE1) were chosen to direct YA into the mitochondrial matrix.

In the case of chloroplasts, several different localizations were targeted within this compartment. Two constructs were made to anchor the YA fusion protein to the chloroplast envelope membranes with the aequorin exposed close to the outer and inner surfaces of the envelope, respectively. To target the aequorin to the outer-envelope surface, a fusion to the full-length outer envelope protein 7 (OEP7; Schleiff et al., 2001) was constructed. OEP7 contains a single N-terminal membrane-spanning α-helix, thus placing the YA outside the organelle. To achieve localization at the inner surface of the inner-envelope, YA was placed behind the first 130 amino acids of Tic40 (At5g16620; Chou et al., 2003). This peptide includes the first membrane-spanning domain of Tic40, which should place the fusion protein at the inner envelope surface. As the final construct, the first 85 amino acids of the chloroplast NADPH-dependent thioredoxin reductase C (NTRC; Perez-Ruiz et al., 2009) were selected to translocate the fusion protein into the chloroplast stroma.

Using the principal constructs as described above, a set of different expression vectors were constructed. Initially, all signal sequences were cloned into the p35S-YA vector and used for transient protoplast transformation. After confirmation of the appropriate localization, the whole expression cassette was transferred into the pBIN and pBINU vectors (Fig. 1), thereby creating two similar sets of binary vectors that just differ in their promoter (Table 1). The UBI10 promoter was included since it is known that expression of foreign proteins driven by the 35S promoter can result in gene silencing (Finnegan and McElroy, 1994; Matzke and Matzke, 1998). In addition, both sets of binary vectors were constructed as two versions each, carrying either a Kanamycin or a Basta (Bar) resistance cassette (Fig. 1) for stable transformation of Arabidopsis TDNA insertion lines or other transgenic lines that already carry one of these resistant markers.

Confirmation of subcellular localization of aequorin fusion proteins

In order to use the different aequorin fusion constructs for the determination of subcellular calcium signals, it is essential to ensure correct localization. In the system described here, this can be achieved by using the fluorescence signal from the YFP included in the fusion constructs (Fig. 1). To establish the proper localization before creation of stable plant lines, all different p35S-YA constructs were transformed transiently into protoplasts from Brassica rapa and localization of the fusion proteins was analysed by laser scanning confocal microscopy (Figs 2, 3). In addition, all constructs were used for the transformation of Nicotiana benthamiana mesophyll cell by Agrobacterium-mediated transformation. Only the results for the mitochondria-targeted construct of the tobacco tranformation are shown representatively. Nevertheless, the localization observed within the two systems was always the same.

Fig. 2.

In vivo localization of different YA fusion proteins in protoplasts from Brassica rapa and Nicotiana benthamiana. The correct localization was confirmed by YFP fluorescence (green). Fluorescence of chlorophyll is shown in magenta. (A) Different YA fusion proteins were expressed in isolated protoplasts from Brassica rapa after PEG-mediated transformation. YFP fluorescence was analysed after 16 h. (B) The mitochondria-localized AKDE1-YA was further analysed by Agrobacterium-mediated transformation of Nicotiana benthamiana leaf cells. Protoplasts were isolated 48 h after transformation and incubated with MitoTracker (blue) before analysis by fluorescence microscopy. The white or black bar within all images represents 20 μm.

Fig. 3.

(A) In vivo localization of chloroplast-targeted YA fusion proteins in protoplasts from Brassica rapa. The correct localization was confirmed by YFP fluorescence (green). Fluorescence of chlorophyll is shown in magenta. Different YA fusion proteins were expressed in isolated protoplasts from Brassica rapa after PEG-mediated transformation. YFP fluorescence was analysed after 16 h. The white or black bar within all images represents 20 μm. (B) Correct localization of the three chloroplast-localized YA variance was further confirmed by Western blot analysis. OE7-YA, Tic40-YA, and NTRC-YA were expressed in Nicotiana tabacum leaf cells by Agrobacterium-mediated transformation. Chloroplasts were isolated 48 h after transformation and analysed with antibodies against aequorin and YFP. Recombinant expressed YA was purified from E. coli and used as a positive control for both antibodies (YA). Thermolysin treatment of isolated chloroplasts confirmed the outer surface exposure of OEP7-YA. Tic40-YA and NRTC-YA were protected from the protease due to their localization within the organelle.

As has been observed before with similar constructs, the fluorescence signal of the YA control containing no targeting sequence could be detected within the cytoplasm as well as in the nucleus (Seibel et al., 2007), indicating that the fusion protein can enter through the nuclear pore (Fig. 2A, first panel). While a cytoplasmic localization excluding the nucleus could be achieved with the CPK17_G2A-NES-YA fusion protein (Fig. 2A, third panel), the fluorescence of the NLS-YA construct was detected exclusively in the nucleus (Fig. 2A, second panel). Thus, these two constructs can be used to distinguish between nuclear and cytoplasmic changes of calcium concentration. The plasma membrane targeted CPK17-YA signal is found solely surrounding the outline of each cell (Fig. 2A, fourth panel) indicating a proper localization within this membrane. Correct targeting of AKDE1-YA to mitochondria was confirmed by the clear overlap between the YFP fluorescence and a MitoTracker Red signal in Brassica rapa (Fig. 2A, fifth panel) as well as in Nicotiana benthamiana protoplasts prepared from leaves after Agrobacterium inoculation (Fig. 2B).

All chloroplast targeted constructs could clearly be observed within the appropriate organelle (Fig. 3). The fluorescence signal of the outer envelope targeted OEP7-YA was found surrounding the red chlorophyll fluorescence indicating a localization in the membrane bordering the organelle (Fig. 3A, upper panel). A more punctate distribution surrounding the red chlorophyll fluorescence was observed for the inner chloroplast membrane targeted Tic40-YA fusion protein (Fig. 3A, second panel). Such a pattern has been observed for other proteins with this localization (Lee et al., 2001; Aseeva et al., 2004; Abdel-Ghany et al., 2005; Duy et al., 2007) thus strongly suggesting the correct targeting of the fusion proteins. By contrast, the NTRC-YA construct targeted to the stroma showed an even distribution within the organelle in accordance with a stromal localization (Fig. 3A, third panel).

The correct localization of the chloroplast targeted fusion proteins was also elucidated by immuno-detection (Fig. 3B). This was done particularly to ensure the correct exposure of the two membrane-associated aequorins on the outer and inner surfaces of the envelope membrane. Chloroplasts were isolated from Nicotiana tabacum leaf cells after Agrobacterium-mediated transformation of each construct and probed with an aequorin as well as a YFP antibody. Proteins of the right size could be detected in each sample (Fig. 3B) indicating that all three fusion proteins are expressed and targeted to the organelle. In addition, intact chloroplasts were treated with thermolysin to remove proteins that extrude from the outer envelope. Figure 3B shows that the OEP7-YA is totally removed by this treatment while the Tic40-YA, as well as the stromal NTRC-YA, was not affected by the protease (Fig. 3B). This suggests that the aequorin and the YFP part of the OEP7-YA fusion are indeed localized on the outer surface of the organelle, while the two other proteins are protected by the envelope membrane.

Plants expressing YA can be used for in vivo calcium measurements

There was a need to confirm that the YFP-aequorin fusion proteins can be used for detecting changes in free subcellular calcium concentration in vivo. As proof of the concept, three stable plant lines were analysed (Fig. 4) expressing the cytosolic (YA), the plasma membrane-associated (CPK17-YA), and the chloroplast outer envelope-associated (OEP7-YA) versions of aequorin. All constructs led to an exposure of the YA fusion protein in the cytoplasm, albeit at different subcellular positions. Since CPK17-YA is localized at the inner surface of the plasma membrane, it should be able to quickly detect signals that are released from extracellular stores. The chloroplast envelope-associated OEP7-YA might be able to detect calcium release from the chloroplast in case this happens in vivo.

Fig. 4.

Cold-induced calcium emission in transgenic Arabidopsis seedlings expressing different variances of cytosol-exposed aequorin. Included in the analysis were CPK17-YA (plasma membrane), OEP7-YA (outer envelope surface), and YA (cytoplasm). (A) Mean values of kinetics obtained from eight different plants in four independent experiments. Plants adapted to an ambient temperature of 20 °C were cooled in two consecutive steps first to 10 °C and then to 0 °C. Light emission was recorded continuously throughout the experiments and the corresponding calcium release calculated is shown in the ordinate. Bars represent the standard deviation at each time point. (B) Average peak value of calcium emission calculated from the first peak of each cooling step. (C) P-values were calculated for the comparison of the different lines with regard to the peak values shown in (B).

Plants were exposed to cold stress by an established set-up using a thermoelectric cooled (Peltier) metal plate (Knight and Knight, 2000). Calcium-induced light emission was detected with a cooled CCD camera system. Four independent experiments were performed and, in each experiment, signals from two different plants were recorded per mutant line. Kinetics obtained from eight different plants in four independent experiments are shown in Fig. 4A, including the standard deviation for each time point. Figure 4B shows averages of the maximum peak heights over all eight plants. In a first step, the plate was cooled from a resting condition of 20 °C to 10 °C and the light emission was recorded for at least 120 s. A clear spike in light emission could be observed in all three lines after a few seconds that lasted for about 20 s (Fig. 4A). Once light emission had gone down to the base-line level, the plate was cooled down to 0 °C and light emission was recorded for another 120 s. Again, a spike in light emission could be observed for all three lines (Fig. 4A). The maximum peak value of the calculated calcium concentration was nearly twice as high as after the first step and also lasted a bit longer. At the end of each measurement, the residual aequorin was discharged, luminescence was measured, and the total released calcium was determined according to previously established methods (Knight and Knight, 2000; Evans et al., 2005).

These results clearly show that all three aequorin constructs are able to detect changes in free calcium concentration. Furthermore, the direct comparison reveals small but significant differences between the three plant lines (Fig. 4C). Plants expressing aequorin associated with the plasma membrane displayed a slightly higher calcium peak value during the first cooling step and the signal appears a few seconds ahead of the signal from the other two lines (Fig. 4A). In the second cooling step, the light emission observed with the soluble aequorin fusion protein is the highest, indicating that internal stores might play a larger role in this reaction. Differences between the signal observed for YA and CPK17-YA are not significant but the OE-YA signal is initially clearly lower compared with the two other lines (Fig. 4B, C). It nevertheless increases in a subsequent second peak to the level observed for the other two lines. Albeit preliminary, these results indicate that differential spatial distribution of calcium release into the cytoplasm can be detected by these different constructs.

Discussion

The use of aequorin is a well-established technique for the detection of cellular calcium in vivo and it has been used previously for the measurement of calcium in the cytosol and a few other plants compartments (Mithofer and Mazars, 2002; Sai and Johnson, 2002). Nevertheless, the binary plasmids used to create these plant lines were designed differently depending on the specific aim of the experiments. By contrast, this work describes a compatible set of binary plasmids based on an equal set-up: all constructs contain an identical YA fusion core placed within an expression cassette comprising promoter, insertion side for the targeting sequence, YA, and terminator (Fig. 1). Targeting sequences are inserted via two unique restrictions sites in front of the YA sequence, thereby allowing an easy expansion of the set with other sequences to target additional compartments.

YFP was included in all fusion constructs to confirm the proper localization of the aequorin reporter. In previously described aequorin lines, confirmation of intra-cellular localization involved laborious fractionation and Western blot analysis or enzyme assays (Knight et al., 1991; Johnson et al., 1995; Mithofer and Mazars, 2002). Using the YA fusion, proper localization can be confirmed easily in transiently transformed protoplasts as well as transgenic plant lines by fluorescence microscopy. In the course of this work, constructs containing different set-ups with regard to the placement of the aequorin and choice of the targeting sequence were tested. Several of these constructs resulted in the mis-localization of the fusion proteins (see Supplementary Table S2 at JXB online). The failure was most often observed for chloroplast-targeted fusion proteins but also for a construct targeted to the mitochondria. Furthermore, our experiments showed that placing NES on the outmost N-terminal end of the fusion protein was not sufficient to exclude the protein from the nucleus (see Supplementary Table S2 at JXB online). While, in most cases, it is unclear which specific feature of these constructs leads to the mis-localization, it clearly shows the importance of this type of monitoring. For all constructs presented in this study, the principal expression as well as the proper localization could be confirmed early on by easy and fast transient expression in both Brassica rapa as well as tobacco. In all subsequently generated transgenic lines of Arabidopsis, a proper localization could be confirmed, indicating the validity of the pre-trials even across species boarders. Thus, inclusion of YFP in the fusion construct will assist in the selection of new constructs in case other targeting sequences are added.

The set of vectors and established plant lines described here also facilitates the easier introduction of aequorin markers into other (mutant) plant lines, i.e. in order to analyse changes in calcium concentration in different genetic backgrounds. Firstly, all constructs are available in two versions, carrying either a Basta or a kanamycin resistance. The choice of two different resistances is important, since mutant plant lines normally already carrying a resistant marker. Even in those cases where both resistances might already be present in the target plants, for example, double TDNA insertion mutants, the selection of lines transformed additionally with the aequorin marker is facilitated by the YFP fusion. Otherwise, the generation of aequorin marker plant lines harbouring multiple T-DNA insertions could require enormous amounts of time and several generations of plants. Propagating plant lines over many generations with 35S driven protein expression can result in reduced expression or even complete gene silencing (Finnegan and McElroy, 1994; Matzke and Matzke, 1998). To speed up the process of selection, the desired binary plasmid can be transformed directly into the target mutant line via the floral dip method or appropriate aequorin-maker lines can be crossed with the target mutant line. To avoid co-suppression of 35S driven protein expression, all constructs are also available with the UBI10 promoter.

Due to the modular set-up of the aequorin plasmids, not only the targeting sequence but also the promoter sequence can easily be exchanged. Further constructs can thus be obtained that would allow the expression of the different aequorin reporters under various promoters, for example, to facilitate expression in certain tissues or at defined developmental stages.

As proof of the concept, calcium release assays were performed after cold shock with three plant lines expressing aequorin soluble in the cytosol as well as attached to the surface of either the plasma membrane or the chloroplast outer envelope. In all lines, spikes of light emission from the aequorin could be observed, indicating that they can be facilitated to measure changes in calcium concentration in vivo. More detailed analysis will have to follow in order to compare the calcium spikes in different aequorin lines. Nevertheless, these preliminary experiments have already indicated that differential spatial distribution of calcium release into the cytoplasm can be detected by these different constructs.

Altogether, these plant lines will serve as valuable tools in future studies of plant calcium signalling, particularly with respect to organelles, which is clearly an emerging field in molecular plant sciences. They will be made available for other researchers upon request.

Abbreviations

Abbreviations
 
• AKDE1

  2-oxoglutarate dehydrogenase subunit E1

 
• CDPK

  calcium-dependent protein kinase

 
• NES

  nuclear export signal

 
• NLS

  nuclear localization signal

 
• NTRC

  NADPH-dependent thioredoxin reductase C

 
• OEP7

  outer envelope protein 7

 
• Tic40

  translocon on the inner chloroplast membrane protein of 40 kDa

 
• YFP

  yellow fluorescence protein

 
• UBI10

  Ubiquitin 10 promoter

 
• YA

  YFP-aequorin

This work was supported by grants from the EU in the Marie-Curie ITN COSI (ITN 2008 GA 215-174) to UCV and MT, by the Deutsche Forschungsgemeinschaft in the ERA-PG project CROPP (VO 656/4-1) to UCV, by the Austrian GEN-AU program in the ERA-PG project CROPP (Project No. 818514) to MT, and by the Austrian Science Foundation (FWF) to MT (P19825-B12) and NM (J3116-B16). MRK would like to thank St John’s College, Durham University for a summer bursary to CHH. MRK and CHH would like to thank Keir Bailey for technical assistance and Dr Heather Knight for scientific advice.