Chloroplast-Specific in Vivo Ca²⁺ Imaging Using Yellow Cameleon Fluorescent Protein Sensors Reveals Organelle-Autonomous Ca²⁺ Signatures in the Stroma

Abstract

In eukaryotes, subcellular compartments such as mitochondria, the endoplasmic reticulum, lysosomes, and vacuoles have the capacity for Ca²⁺ transport across their membranes to modulate the activity of compartmentalized enzymes or to convey specific cellular signaling events. In plants, it has been suggested that chloroplasts also display Ca²⁺ regulation. So far, monitoring of stromal Ca²⁺ dynamics in vivo has exclusively relied on using the luminescent Ca²⁺ probe aequorin. However, this technique is limited in resolution and can only provide a readout averaged over chloroplast populations from different cells and tissues. Here, we present a toolkit of Arabidopsis (Arabidopsis thaliana) Ca²⁺ sensor lines expressing plastid-targeted FRET-based Yellow Cameleon (YC) sensors. We demonstrate that the probes reliably report in vivo Ca²⁺ dynamics in the stroma of root plastids in response to extracellular ATP and of leaf mesophyll and guard cell chloroplasts during light-to-low-intensity blue light illumination transition. Applying YC sensing of stromal Ca²⁺ dynamics to single chloroplasts, we confirm findings of gradual, sustained stromal Ca²⁺ increases at the tissue level after light-to-low-intensity blue light illumination transitions, but monitor transient Ca²⁺ spiking as a distinct and previously unknown component of stromal Ca²⁺ signatures. Spiking was dependent on the availability of cytosolic Ca²⁺ but not synchronized between the chloroplasts of a cell. In contrast, the gradual sustained Ca²⁺ increase occurred independent of cytosolic Ca²⁺, suggesting intraorganellar Ca²⁺ release. We demonstrate the capacity of the YC sensor toolkit to identify novel, fundamental facets of chloroplast Ca²⁺ dynamics and to refine the understanding of plastidial Ca²⁺ regulation.

Chloroplasts are defining organelles of photosynthetic eukaryotes. In addition to photosynthesis, plastids accommodate several essential pathways, such as fatty acid biosynthesis, making them indispensable, even in nonphotosynthetic tissues where they occur as nongreen plastid types (Neuhaus and Emes, 2000). Initial observations of Ca²⁺ accumulation in chloroplasts date back over 75 years (Neish, 1939), and more recent reports have shown that Ca²⁺ is primarily present in the thylakoid lumen where free concentrations can reach the millimolar range (Yamagishi et al., 1981; Brand and Becker, 1984). In vivo monitoring of stromal free Ca²⁺ concentration ([Ca²⁺]_str) using the aequorin probe subsequently enabled the observation of diurnal fluctuations in [Ca²⁺]_str between the high nanomolar and the low micromolar range (Johnson et al., 1995; Sai and Johnson, 2002; Nomura et al., 2012). The transition from light to darkness, in particular, was shown to induce stromal Ca²⁺ accumulation (Johnson et al., 1995; Sai and Johnson, 2002; Sello et al., 2016). Consequently it has been hypothesized that chloroplasts sequester Ca²⁺ into the thylakoids during the day and release it into the stroma at onset of darkness (Sai and Johnson, 2002). This rise of [Ca²⁺]_str has been suggested to act as posttranslational regulator inhibiting photosynthetic CO₂ fixation (Portis and Heldt, 1976; Demmig and Gimmler, 1979) based on the observation that the Calvin cycle enzyme Fru-1,6-bisphosphatase is inhibited by high concentrations of Ca²⁺ in vitro (i.e. 500 μm for spinach [Spinacia oleracea]; Charles and Halliwell, 1980; Hertig and Wolosiuk, 1983; Kreimer et al., 1988). Ca²⁺ regulation of chloroplast protein translocation is mediated by calmodulin binding to Tic32, a member of the TIC translocon. This calmodulin regulation appears to occur from the stromal side of the chloroplast envelope consistent with direct function of stromal Ca²⁺ regulation for regulation of protein import in individual plastids (Chigri et al., 2006). As a particularly intriguing component of the chloroplast Ca²⁺ regulatory network, the calcium sensing protein CAS, which is localized in chloroplast thylakoid membranes, has been associated with a range of regulatory functions such as guard cell dynamics, photoacclimation, and immune responses (Nomura et al., 2008; Vainonen et al., 2008; Weinl et al., 2008; Petroutsos et al., 2011).

Despite quantitative estimation and measurement of relative changes in free Ca²⁺ concentrations in different chloroplast subcompartments using aequorin (Mehlmer et al., 2012; Sello et al., 2016), the extent and direction of intraplastidial Ca²⁺ fluxes or Ca²⁺ transport into and from the cytosol are currently not understood and none of the transporters involved has been identified (Rocha and Vothknecht, 2012; Finazzi et al., 2015). Accordingly, it remains unknown how stromal Ca²⁺ dynamics and signaling are orchestrated. This is a particularly pressing question considering the hypothetical contribution from two Ca²⁺ sources, from the outside (cytosol) and from within (thylakoid lumen). Such a concept is unique for an endosymbiotic organelle and allows the integration of information from both the former host as well as the former endosymbiont. This relationship is likely to include tissue-, cell-, and organelle-specific effects that make it highly desirable to perform Ca²⁺ measurements with high spatiotemporal resolution. However, aequorin-based luminescence measurements that currently form the basis of our understanding of the in vivo dynamics of free Ca²⁺ dynamics in plants are limited with respect to the spatial resolution (Knight et al., 1991, 1996). Fluorescent Förster resonance energy transfer (FRET)-based Ca²⁺ reporter proteins match all requirements for specific Ca²⁺ analyses in chloroplasts with tissue, cell, or organelle resolution (Choi et al., 2012, Costa and Kudla, 2015). Despite significant efforts by several laboratories, the generation of such sensor lines has proved difficult, however, and their lack has hampered gaining further insight into chloroplastic Ca²⁺ dynamics in vivo.

Here, we report the establishment of chloroplastic Yellow Cameleon (YC) lines in Arabidopsis (Arabidopsis thaliana) as a toolset to explore the unique characteristics of stromal Ca²⁺ dynamics. We demonstrate the functionality of the sensing approach in vivo in plastids of roots and leaves and show that at tissue level the measured stromal Ca²⁺ signatures resemble those reported using aequorin (Sello et al., 2016). In addition, we report transient Ca²⁺ in single organelles, as a hitherto unknown component of stromal Ca²⁺ dynamics at the level of the single chloroplast, demonstrating the added power of the fluorescent sensor setup. Further dissecting the Ca²⁺ signatures induced by light-to-low-intensity blue light illumination transitions, we present evidence for a gradual Ca²⁺ mobilization from intrachloroplastic stores, in conjunction with stromal Ca²⁺ spikes that appears to depend on the cytosol as a Ca²⁺ source but can be modulated at the level of the individual chloroplast.

RESULTS

Specific Targeting of Cameleon Probes to the Plastid Stroma

In order to analyze the in vivo Ca²⁺ dynamics in the chloroplast stroma at high resolution and specificity, we aimed to establish a robust sensing system. We selected two variants of the YC probes (Miyawaki et al., 1997) as genetically encoded fluorescent sensors covering different concentration ranges of free Ca²⁺: YC3.6 with a single in vitro K  _d value of 250 nm and YC4.6 with two in vitro K  _d values at 56 nm and 14 µm (Nagai et al., 2004). YC3.6 has been used successfully to monitor cytosolic, nuclear, and mitochondrial Ca²⁺ dynamics in plant cells (Krebs et al., 2012; Loro et al., 2012; Loro and Costa, 2013), whereas the YC4.6 sensor has been used to monitor the Ca²⁺ dynamics in the endoplasmic reticulum (ER) of Arabidopsis pollen tubes (Iwano et al., 2009). We reasoned that different transgenic lines expressing probes with different Ca²⁺ affinities would cover a wide range of free Ca²⁺ concentrations. The lower K  _d of the YC4.6 sensor (56 nm) would allow the detection of small [Ca²⁺]_str transients, similar to those reported for the cytosol using the Nano-Cameleon sensor (in vitro K  _d at 65 nm; Choi et al., 2014b). Pronounced [Ca²⁺]_str transients peaking in the low to medium micromolar concentration range would be measureable based on the second K  _d (14 µm), with the intermediate range covered by YC3.6. In order to express both sensors in the stroma, we chose the targeting peptide of stromal β-amylase 4 from Arabidopsis (Bam4, AT5G55700; Fulton et al., 2008). We first focused on YC3.6 to obtain proof-of-concept for sensor targeting. We fused the coding sequence for the first 62 amino acids of Bam4 to the N terminus of the YC3.6 sequence. The fusion construct was placed into a 0029 pGreen backbone under the control of a single CaMV35S promoter (Hellens et al., 2000; Bonza et al., 2013). To verify subcellular localization, we transiently expressed the construct in tobacco (Nicotiana tabacum) leaves and analyzed the cpVenus (an improved YFP version present in the Cameleon YC3.6; Nagai et al., 2004) fluorescence by confocal laser scanning microscopy (CLSM). The probe was targeted to chloroplasts; however, a considerable proportion of cpVenus fluorescence was observed in the cytosol and the nucleus (Supplemental Fig. S1, A–C). Based on similar observations in mammalian mitochondria (Filippin et al., 2005), we reasoned that the single Bam4 targeting sequence was not sufficient for efficient plastid targeting and aimed to improve the stromal localization by adding the targeting peptide in tandem leading to the generation of the pGreen 0029 35S:2Bam4-YC3.6 construct. Expression of the modified construct in tobacco leaves revealed complete plastid localization of the YC3.6 probe (Supplemental Fig. S1, D–F), without detectable cpVenus signal in the cytosol or the nucleus (Supplemental Fig. S1, D and F). The same was confirmed in the subsequent Arabidopsis work (see next section and Supplemental Fig. S11). Hence, complete plastid localization of the YC3.6 sensor could be achieved by fusion to a tandem targeting sequence, allowing unambiguous analyses in the chloroplast stroma.

Overcoming Sensor Silencing to Generate Yellow Cameleon Sensor Lines for Ca²⁺ Sensing in the Plastid Stroma in Arabidopsis

We generated stable sensor lines by transforming Arabidopsis ecotype Columbia-0 (Col-0) with the 2Bam4-YC3.6 construct. However, despite the selection of four independent transgenic lines based on the kanamycin resistance marker, no fluorescence was detected in leaves of young seedlings. Sensor fluorescence was only detected in the columella cells and in individual root epidermal cells of 4-d-old seedlings, while practically no fluorescence could be detected in the root of older seedlings (12 d old; Supplemental Fig. S2). This was indicative of strong transgene silencing, which we and others have already observed in other fluorescent protein lines (Deuschle et al., 2006; Jones et al., 2014; Behera et al., 2015; Schwarzländer et al., 2016). To overcome this limitation, we introduced the 35S:2Bam4-YC3.6 construct into the Arabidopsis rdr6-11 background (here referred to as rdr6 for simplicity), which is impaired in gene silencing (Peragine et al., 2004). Rdr6 was successfully employed to overcome problems related to transgene silencing upon expression of Glc and ABA nanosensors (Deuschle et al., 2006; Jones et al., 2014). We transformed rdr6 plants with the 35S:2Bam4-YC3.6 construct and 12 independent transformants were isolated by antibiotic selection. Confocal microscope analyses of cotyledon cells (Fig. 1, A–C), seedling root cells (Fig. 1, D–I), and cells in leaves of mature plants (Fig. 1, L–N) detected intense and uniform cpVenus fluorescence across all tissues. The fluorescent signal colocalized with chlorophyll autofluorescence in cotyledon and leaf epidermal cells (Fig. 1, C and N; Supplemental Fig. S3), confirming plastid localization of the 2Bam-YC3.6 probe. We also adopted this expression strategy for the YC4.6 sensor, isolated four independent transgenic lines, for which suitable expression and plastidial localization were confirmed across tissues and developmental stages (Supplemental Fig. S4, A–N). In parallel, we generated rdr6 plants expressing the YC3.6 sensor in the cytosol for reference, using a construct containing an N-terminal nuclear export signal (NES-YC3.6; Krebs et al., 2012). We selected six independent transgenic lines and CLSM imaging confirmed exclusive localization to the cytosol (Supplemental Fig. S5). Stimulus-triggered Ca²⁺ signatures were identical in the wild type and rdr6 background expressing the NES-YC3.6 sensor (see next section), suggesting that Ca²⁺ responses in measured in rdr6 were indeed representative.

Figure 1.

Expression of 2Bam4-YC3.6 in Arabidopsis rdr6. Sensor expression was analyzed in cotyledons (A–C), upper roots (E and F), and root tips (G–I) of 4-d-old seedlings and in leaves of mature plants (L–N). A, D, G, and J, cpVenus fluorescence in green. B and K, Chlorophyll autofluorescence (Chl). E and H, Bright field. C, F, I, and L, Overlay (Ol) of the two channels shown for each tissue. Absence of strict colocalization between chlorophyll and cpVenus signal for the chloroplasts is accounted for by the localization of the sensor in the stroma rather than the thylakoids emitting the chlorophyll signal and nongreen plastids from the epidermal pavement layer. The exemplary images shown are representative of the cpVenus fluorescence of Cameleon being present in all tissues, suggesting absence of silencing.

Cameleon Lines Allow Faithful in Vivo Ca²⁺ Monitoring in the Plastid Stroma of Roots and Unravel Linkage to Cytosolic Ca²⁺ Transients

To investigate the functionality and reliability of the Ca²⁺ probes in vivo, we performed a series of Ca²⁺ imaging analyses. Initially we focused on nongreen plastids in roots, which allow efficient and reproducible application of chemical stimuli (Loro et al., 2012; Bonza et al., 2013; Behera and Kudla, 2013, Behera et al., 2015). Extracellular ATP (eATP) acts as a signaling molecule involved in plant development and stress responses (Cao et al., 2014; Tanaka et al., 2014; Choi et al., 2014a). eATP administration to root tip cells has been reported to trigger a cytosolic Ca²⁺ transient in both aequorin expressing seedlings (Tanaka et al., 2010) and Cameleon (Tanaka et al., 2010; Krebs et al., 2012; Loro et al., 2012) with a consequent Ca²⁺ accumulation in intracellular compartments including the mitochondria and the ER (Loro et al., 2012; Bonza et al., 2013; Wagner et al., 2015b). Specifically, in the root tip, eATP is initially perceived by the cells of the meristematic zone (lateral root cap cells), and afterward a Ca²⁺ wave travels from this region to the elongation zone (Tanaka et al., 2010; Loro et al., 2012; Costa et al., 2013). We therefore analyzed plastidial Ca²⁺ dynamics in response to 500 µm  eATP in the root meristematic cells of seedlings. In parallel, we monitored the cytosolic Ca²⁺ dynamics using the NES-YC3.6 line as a control (Krebs et al., 2012).

eATP induced a rapid Ca²⁺ transient in the cytosol of rdr6 seedlings (Fig. 2A, upper row), as observed before in the wild-type background (Loro et al., 2012, Bonza et al., 2013; this work), which coincided with transient Ca²⁺ accumulation in the plastid stroma (Fig. 2, A–C). Both probes efficiently reported the accumulation of Ca²⁺ in the plastid stroma with similar kinetics (red and blue traces of FRET ratios in Fig. 2B and normalized ratio changes in Fig. 2C). For both probes, the dynamics were caused by a bona fide FRET response (Supplemental Fig. S6, A and B, illustrating the increase in cpVenus and the decrease in CFP fluorescence intensity) and the setup allowed measurement of the transients in a genuine in vivo situation (Fig. 2B–2G). However, the characteristics of the Ca²⁺ transients differed markedly between both compartments. Interestingly, the basal steady-state FRET ratio was consistently lower in the stroma as compared to the cytosol, indicating lower free Ca²⁺ concentrations. Our calibration estimates 86 ± 9 nm of free Ca²⁺ in plastid stroma as compared to 108 nm ± 8 in the cytosol (Fig. 2H; see Supplemental Discussion). This is opposite to the mitochondrial matrix where YC3.6 consistently indicates higher basal Ca²⁺ levels (Wagner et al., 2015b) and in line with what has been observed for the relative concentrations of stromal versus cytosolic Ca²⁺ in aequorin-based measurements (Sai and Johnson, 2002). Furthermore, the transient amplitude was more pronounced in the cytosol than in the plastid stroma, and the maximum 2Bam4-YC3.6 FRET-ratio of the stroma reached only slightly above the corresponding steady-state FRET ratio of the cytosol (Fig. 2, B and E). The rate of plastidial Ca²⁺ increase was also lower in comparison to that of the cytosol (Fig. 2, B and F). This mirrors recent aequorin-based observations where different stimuli (NaCl, mannitol, cold shock, and oligogalacturonides) applied to cultured Arabidopsis cells induced stromal Ca²⁺ elevations that were lower than those in the cytosol (Sello et al., 2016). Both stromal Cameleon probes indicated the maximum ratio change at about 140 s after the eATP stimulus (150 ± 13.5 s and 136 ± 19.9 s for 2Bam-YC3.6 and 2Bam4-YC4.6 lines respectively; Fig. 2F) while the maximum was already reached after about 40 s in the cytosol (40 ± 11.5 s). Moreover, the recovery of the resting prestimulus FRET ratio in the cytosol was more rapid than in plastids (Fig. 2G). The cytosolic FRET ratio response was halved after 96.5 ± 20.3 s, while the stromal FRET ratios required 300 ± 70 s for the 2Bam4-YC3.6 and 136.2 ± 10.3 for the 2Bam4-YC4.6, respectively. These results demonstrate that both probes report Ca²⁺ dynamics in the plastid stroma with similar response characteristics that are consistent with those that have been reported using aequorin. The 2Bam4-YC3.6 sensor exhibited a more pronounced FRET response than 2Bam4-YC4.6 (amplitude; Fig. 2, B, C, and E), which are likely to reflect the differential Ca²⁺ binding affinities of both sensors. Hence, our observations empirically identify YC3.6 as the sensor with the preferable characteristics to monitor physiological Ca²⁺ dynamics in the plastid stroma, consistent with previous aequorin-based estimations of the free Ca²⁺ concentrations in the stroma in the high nanomolar to low micromolar range (Johnson et al., 1995; Sello et al., 2016).

Figure 2.

Plastidial and cytosolic Ca²⁺ dynamics in root tips treated with eATP. A, Ratiometric false-color images from representative time series of rdr6 root tips expressing NES-YC3.6 (cytosolic, first row), 2Bam4-YC3.6 (plastidial, second row), and 2Bam4-YC4.6 (plastidial, third row) at treatment with 500 µm eATP (n ≥ 4). The number in the images indicates the time passed after acquisition start in seconds. The black line below the images indicates the duration of eATP exposure (for 150 s followed by washout). B and C, the region of interest in the meristematic/lower elongation zone as indicated by white circles was analyzed and plotted over time for the averaged FRET ratio ± sd (B) and for the normalized FRET ratio, as ƊR/R₀ ± sd (C). The black line under the graphs indicates the treatment with eATP corresponding to A. D, Steady-state FRET ratios preceding eATP application (averaged over a 45-s time window before eATP application each). E, Maximal relative amplitude of FRET ratio triggered by eATP application. F, Time to reach maximal amplitude of FRET ratio after start of eATP application. G, Time to pass half-maximal FRET ratio measured after reaching the maximal relative amplitude of FRET shown in E. H, Transformation of FRET ratio data of NES-YC3.6 and 2Bam4-YC3.6 shown in C to absolute free Ca²⁺ concentration, using the calibration rationale described in the Supplemental Discussion. I, FRET ratios of NES-YC3.6 and 2Bam4-YC3.6 in root tips at treatment with 500 µm eATP in presence of 1 mm EGTA or 10 mm of CaCl₂. n ≥ 4; error bars = sd; *P ≤ 0.01, **P ≤ 0.05, °P ≥ 0.1 (t test).

To further dissect the link between the eATP-triggered plastidial and the cytosolic Ca²⁺ transients, we evaluated the dose dependence of the response using 10, 100, and 500 µm of eATP (Supplemental Fig. S7, A–D). To assess for potential effects of the rdr6 background, we also evaluated the cytosolic Ca²⁺ responses to eATP in the wild-type NES-YC3.6 line (Supplemental Fig. S7, E and F). eATP triggered stromal and cytosolic Ca²⁺ transients where the amplitude correlated with the concentration of the stimulus (Supplemental Fig. S7, A–F). The higher responsiveness of the 2Bam4-YC3.6 sensor compared to the 2Bam4-YC4.6 sensor was confirmed with the difference decreasing at lower stimulus strength consistent with the lower K  _d of YC4.6 (about 56 nm) allowing in vivo sensing close to resting levels of free Ca²⁺ in the stroma (reported at about 100–150 nm [Nomura et al., 2012; Stael et al., 2012] and 86 ± 9 nm from our estimation; Fig. 2H; Supplemental Fig. S7C). The cytosolic Ca²⁺ responses determined by NES-YC3.6 in rdr6 and wild-type background were indistinguishable, indicating no major impact of the rdr6 background on cellular Ca²⁺ regulation. The temporal correlation of the transients in the stroma and the cytosol as well as the clear dose-dependent response to eATP in both compartments suggest that cytosolic and plastidic Ca²⁺ signals are linked.

To address this further, we repeated the experiment shown in Figure 2B in the presence of 1 mm EGTA to chelate extracellular Ca²⁺. Both cytosolic and plastidial Ca²⁺ elevation were nearly completely abolished (Fig. 2I) but were reestablished after addition of 10 mm Ca²⁺ followed by another eATP stimulus (Fig. 2I). This result demonstrates that stromal Ca²⁺ elevation is secondary to that in the cytosol and likely generated by Ca²⁺ uptake from the cytosol across the plastid envelope predicting an efficient Ca²⁺ uptake system. While the identity of components involved remains to be determined, the differential response characteristics between both compartments point to significant integration at the level of Ca²⁺ uptake (Wagner et al., 2016).

Chlorophyll Autofluorescence Does Not Affect Cameleon Measurements in Chloroplasts

We next aimed to assess the feasibility of stromal free Ca²⁺ measurements by Cameleon in green plastids. However, a general concern associated with using fluorescent sensors in chloroplasts is potential spectral interference from the photosynthetic pigments leading to measurement artifacts. To test whether this affects the measured cpVenus/CFP FRET ratio in vivo, we exploited the resolution of CLSM imaging to compare the YC3.6 FRET ratios determined from the thylakoid containing chloroplast centers, where chlorophyll autofluorescence was present (Fig. 3A), with those from stromules that show sensor fluorescence but no chlorophyll (Fig. 3A; Osteryoung and Pyke, 2014) in Arabidopsis hypocotyls. FRET ratios were indistinguishable between both regions of interest (Fig. 3B), indicating that photosynthetic pigments do not spectroscopically interfere with the sensor readout to any significant extent.

Figure 3.

Impact of chlorophyll fluorescence on YC3.6 FRET ratio. A, Hypocotyl of Arabidopsis seedlings expressing 2Bam4-YC3.6 imaged by CLSM. cpVenus signal, green; chlorophyll fluorescence, magenta; cpVenus/CFP ratio, ratiometric false-color scale. Regions of interest were defined for chloroplast centers where chlorophyll signal was present (straight lines) and stromules where no chlorophyll signal was detected (dotted lines). B, Resulting FRET ratios (cpVenus/CFP) for chloroplast stromules and centers. n = 23; error bars = sd; ns P > 0.25 (t test).

Stromal Ca²⁺ Elevations after Light-to-Low-Intensity Blue Light Illumination Transition in Chloroplasts

As a stimulus to induce Ca²⁺ dynamics in chloroplasts we chose the transition from light to low-intensity blue light to mimic a situation similar to darkness that had been previously observed to trigger Ca²⁺ elevations using stromal-targeted aequorin at the whole plant and cell culture level (Johnson et al., 1995; Sai and Johnson, 2002; Sello et al., 2016). Sunset and variable shading represent processes reflecting the physiological significance associated with this stimulus. We imaged cotyledon mesophyll cells of intact seedlings resolving individual chloroplasts (Fig. 4A, see arrows; Supplemental Movie S1). The seedlings were illuminated with white light (100 µE m⁻² s⁻¹) until the start of the measurement, which represented the onset of low-intensity blue light phase (0 s in the graph). In our setup, we consider low-intensity blue light illumination the situation in which the chloroplasts receive light only from the fluorescent pulsed lamp flashes (at 436 ± 10 nm, 100 µE m⁻² s⁻¹ with the 20× magnification) that are strictly required for microscopy. A gradual increase in stromal free Ca²⁺ was evident based on a change in FRET ratio, evidenced by counterparallel behavior of CFP and cpVenus fluorescence (Fig. 4B). No increase of plastidial Ca²⁺ concentration was observed when the same regime was applied to root plastids (Supplemental Fig. S8), indicating that the effect was specific to chloroplasts. Recent work that was published during the preparation of this article showed consistent behavior between green and nongreen cell cultures (chloroplasts versus amyloplasts), confirming our observation of differences in the Ca²⁺ responsiveness between plastid types (Sello et al., 2016). These observations further confirm that fluorescent Ca²⁺ sensing in vivo provides results that are generally consistent with the responses measured with aequorin at tissue level, but with resolution down to the individual organelle. To validate our technical protocol for Cameleon FRET analysis under light-to-low-intensity blue light illumination transitions, we assessed the photosynthetic activity of the chloroplasts under those conditions. Leaves of Arabidopsis wild type were imaged by pulse amplitude modulation (PAM) imaging (Schreiber, 1986; Supplemental Fig. S9 and Supplemental Discussion). We used 400 ms of illumination at 2000 µE m⁻² s⁻¹ of red light (660 nm) repeated every 5 s. Such conditions only partially reproduce the microscope situation. Because of the technical limitations of the PAM imaging, we could not use 100 s µE m⁻² s⁻¹. Nevertheless, we observed a pronounced drop in photosynthetic activity when the described regime was applied to nearly zero (from 1 to 0.12 1-qP; Supplemental Fig. S9), confirming that our low-intensity blue light illumination regime carries physiological meaning.

Figure 4.

2Bam4-YC3.6 reports stromal Ca²⁺ accumulation at light-to-low-intensity blue light illumination transition. A, Ratiometric false-color images at 100-s intervals from a representative time series of the 2Bam4-YC3.6 response to a light-to-low-intensity blue light illumination transition in an Arabidopsis rdr6 cotyledon (n = 8). The number in the images indicates the time in seconds passed since onset of low-intensity blue light illumination. B, cpVenus and CFP fluorescence and normalized FRET ratio (cpVenus/CFP) as ƊR/R₀, corresponding to the region of interest indicated by the red box in the first ratiometric image of A. The start of the measurements coincided with the onset of low-intensity blue light illumination as indicated by the white bar above the graphs (light) and the black bar (low-intensity blue light illumination). The gray panels show a close up of cpVenus and CFP fluorescence that highlight the FRET response. Fluorescence A.U., arbitrary units.

Single Chloroplast Analyses Reveal Short-Duration Spiking as a Component of Stromal Ca²⁺ Dynamics

Measurements averaged across large, heterogeneous populations of cells or organelles generate robust data at the cost of losing information about the individual cell or organelle (Dodd et al., 2006). To overcome this limitation, functional microimaging approaches in vivo are particularly suitable and have contributed to the discovery of localized Ca²⁺ sparks (Cheng et al., 1993) and microdomains (Rizzuto and Pozzan, 2006) as well as bioenergetics transients of individual mitochondria (Schwarzländer et al., 2012; Breckwoldt et al., 2014). We therefore next analyzed the Ca²⁺ dynamics of single chloroplasts in mesophyll cells from entire cotyledons by CLSM (Fig. 5, A and B), and mature leaf stomata guard cells by wide-field microscopy (Fig. 5, C–E).

Figure 5.

Stromal Ca²⁺ spiking and Ca²⁺ accumulation at light-to-low-intensity blue light illumination transition. A, Cotyledon mesophyll cells of Arabidopsis rdr6 seedlings expressing 2Bam4-YC3.6 imaged by CLSM. Representative spiking events of the sensor observed in individual plastids. CpVenus signal, green; CFP signal, magenta. Bar = 5 µm. B, Quantitative analysis of FRET ratios of the chloroplasts shown in A. Intensities of cpVenus (green) and CFP (magenta) are shown as insets in the cpVenus/CFP FRET ratio graphs. C, Ratiometric false-color images from representative time series of guard cell chloroplasts expressing 2Bam4-YC3.6 imaged by a wide-field microscope in response to light-to-low-intensity blue light illumination transition (n = 15). The number in the images indicates the time in seconds passed since onset of low-intensity blue light illumination. D, Ratio of cpVenus/CFP corresponding to the regions of interest indicated by the numbers in the first ratiometric image of C. E, Detail of D to illustrate spiking of individual chloroplasts, as a representative time series. The start of the measurements coincided with the onset of low-intensity blue light illumination as indicated by the white bar above the graphs (light) and the black bar (low-intensity blue light illumination).

The analysis of single mesophyll chloroplasts (n = 20) revealed that single chloroplasts showed infrequent stromal Ca²⁺ transients that can occur in isolation but also repetitively (Supplemental Fig. S10 and Supplemental Discussion). The transients are the result of bona fide FRET between cpVenus and CFP of the 2Bam4-YC3.6 probe (Fig. 5B, insets), strongly suggesting true Ca²⁺ spiking. Moreover, the spikes were limited to the individual chloroplasts, and no comparable transients were observed in the cytosol (Supplemental Fig. S11). Additionally, in general, no discernable synchronization with other chloroplasts of the same cell was apparent (Supplemental Fig. S11), with the exception of chloroplast pairs (Fig. 5A). Consistent with our observations at the level of entire cotyledons, we observed that shortly after the light-to-low-intensity blue light illumination switch, an increase in [Ca²⁺]_str was detected in individual chloroplasts. This elevation typically started 3 min after onset of low-intensity blue light illumination and reached a plateau after increasing for about 5 min. At the level of the single chloroplast, the increase appeared faster, leading to higher Ca²⁺ concentration amplitudes in comparison to the measurements that average out over chloroplast populations in cotyledon mesophyll. This is likely to illustrate the impact of averaging across heterogeneous plastid populations of a leaf and also to mixing with nongreen plastids in the epidermis. Yet, both analyses showed sustained low-intensity blue light illumination-induced stromal free Ca²⁺ elevation, with a slow recovery phase that exceeded 1 h (Fig. 5D). The recovery was a bona fide stromal Ca²⁺ decrease since the FRET response showed inverse behavior of the fluorescence intensities CFP and cpVenus (Supplemental Fig. S12). A statistical analysis revealed that a sustained Ca²⁺ increase occurred in 92% of guard cell chloroplasts (n = 87). We also observed Ca²⁺ spiking overlaying the sustained Ca²⁺ elevation, suggesting that it represents a distinct component of stromal Ca²⁺ dynamics (Fig. 5, D and E). The spikes were not detectable at the level of larger cotyledon tissue areas which, on average, showed a smooth response curve, similar to Figure 4, A and B, where analysis was performed across chloroplast populations similar to previous observations using aequorin as a Ca²⁺ reporter (Dodd et al., 2006). Ca²⁺ spikes were detectable in 83% of guard cell chloroplasts (n = 65) and were clearly distinguishable from the background noise as short-lived and reversible Ca²⁺ elevations (as exemplified by the traces of Fig. 5E and in the Supplemental Movie S2). No spikes were observed after stromal free Ca²⁺ reached the steady state values after about 500 s at low-intensity blue light illumination (Fig. 5, C and D), indicating that the underlying fluxes do not occur, or cannot be resolved, at elevated Ca²⁺ concentration. While the gradual and sustained Ca²⁺ influx showed similar dynamics across the individual guard cell chloroplasts, the spikes were not temporally synchronized between individual chloroplasts of the same cell (as observed for mesophyll cells; exemplified by the colored traces in Fig. 5, C–E). The unique spiking pattern for each chloroplast strongly suggests that these Ca²⁺ signals can be modulated at the level of the single organelle.

Stromal Sustained Increases and Spikes of Ca²⁺ Concentrations Depend on Distinct Ca²⁺ Sources

To elucidate the origin of both components of stromal Ca²⁺ dynamics that can be observed during light-to-low-intensity blue light illumination transitions, we investigated if and how they were connected to guard cell Ca²⁺ dynamics in the cytosol. We detected cytosolic Ca²⁺ oscillations (Supplemental Fig. S13, A and B), consistent with previous reports in which the frequency and amplitude of cytosolic Ca²⁺ oscillations were directly linked to stomatal aperture (Allen et al., 2000; Yang et al., 2008; Kudla et al., 2010). To modulate cytosolic Ca²⁺ dynamics in the guard cells, we exploited their dependency on the availability and concentration of apoplastic Ca²⁺. Chelating Ca²⁺ of the extracellular stores by EGTA resulted in complete loss of the cytosolic oscillations (Supplemental Fig. S12, A and B), while the basal cytosolic FRET ratio was not affected indicating that steady-state Ca²⁺ levels were not affected (Supplemental Fig. S13, C–E). In contrast, depleting external Ca²⁺ stores did not modify stromal Ca²⁺ signatures within chloroplasts. Both the gradual, sustained free Ca²⁺ accumulation as well as Ca²⁺ spiking were maintained (Fig. 6, A and B), and the percentage of chloroplasts showing Ca²⁺ spiking in presence (86%; n = 35) or absence of EGTA (83%; n = 65) was not distinguishable. These results indicate that stromal Ca²⁺ dynamics at the light-to-low-intensity blue light illumination transition do not mirror nor depend on cytosolic Ca²⁺ dynamics and suggest a large degree of autonomy in chloroplast Ca²⁺ handling. However, this does not exclude the possibility that the cytosol serves as source for chloroplast Ca²⁺ accumulation, considering that EGTA-mediated depletion of extracellular Ca²⁺ stores did not impact on basal Ca²⁺ levels in the cytosol (Supplemental Fig. S13E). To examine this in more depth, we performed the EGTA treatment in the presence of the permeabilizing detergent digitonin as previously reported in Costa et al. (2010). The effectiveness of permeabilization was confirmed by time-lapse experiments performed with guard cells of NES-YC3.6 plants showing that both the addition and removal of Ca²⁺ were efficiently reported by the cytosolic Cameleon sensor (Supplemental Fig. S13). The chelation of extrachloroplastic Ca²⁺ did not affect the basal stromal Ca²⁺ concentration (Fig. 6D) nor the gradual, sustained Ca²⁺ increase induced by the light-to-low-intensity blue light illumination transition, but had a pronounced impact on stromal Ca²⁺ spiking (Fig. 6C). The percentage of chloroplasts in which Ca²⁺ spiking was observed drastically decreased from 86 to 15% (n = 40; Fig. 6C), suggesting that stromal Ca²⁺ spiking depends on extrachloroplastic Ca²⁺ and/or cellular integrity, while the gradual Ca²⁺ elevation completely relies on intrachloroplastic stores. These results identify two distinct sources for the two different components that shape stromal Ca²⁺ dynamics at light-to-low-intensity blue light illumination transitions and support the existence of distinct Ca²⁺ signaling processes that are integrated in the stroma of chloroplasts.

Figure 6.

Components of stromal Ca²⁺ dynamics in single chloroplasts at light-to-low-intensity blue light illumination transition. A, Normalized FRET ratios (cpVenus/CFP), as ƊR/R₀ from 10 different single chloroplasts (gray lines) in guard cells during light-to-low-intensity blue light illumination transition under control conditions (CNT). The red line represents the average of the 10 individual traces/chloroplasts. B, Same as A, but in the presence of 1 mm EGTA. C, Same as A, but in permeabilized cells (500 μm digitonin) and in the presence of 1 mm EGTA (Perm+EGTA). The start of the measurements coincided with the onset of low-intensity blue light illumination as indicated by the white bar above the graphs (light) and the black bar (low-intensity blue light illumination). D, Percentage of chloroplasts that showed oscillations under the conditions shown in A to C and within the observation time of 900-s CNT, 54 chloroplasts (65 total); EGTA, 30 chloroplasts (35 total); Perm+EGTA, 6 chloroplasts (40 total); §P < 0.001, ns > 0.5 (χ² test). E, Comparison of the resting stromal Ca²⁺ levels and maximum Ca²⁺ levels after light-to-low-intensity blue light illumination transition reported as FRET ratio of cpVenus/CFP ± sd; *P ≤ 0.01, ns P > 0.25 (t test).

DISCUSSION

In this study, we introduce a toolkit for the exploration and dissection of Ca²⁺ dynamics in the plastid stroma of Arabidopsis. This includes transgenic lines expressing two plastid-localized Cameleon sensors enabling the investigation of different concentration ranges of free Ca²⁺, (2Bam4-YC3.6 and 2Bam4-YC4.6), in addition to the cytosolic-localized YC3.6 sensor (NES-YC3.6) in the silencing-suppressed rdr6 background, allowing specific and reliable sensor expression. When comparing cytosolic sensor responses in rdr6 and wild-type background, we did not detect any difference in Ca²⁺ responses. Although it is impossible to assess all potential stimulus/compartment combinations, it appears conclusive to deduce that Ca²⁺ measurements made in the rdr6 background faithfully reflect the situation in the wild type. We demonstrate that both Cameleon probes react dynamically in vivo and, hence, cover a suitable dynamic range of free Ca²⁺ (as set by sensor K  _d) to facilitate measurements of Ca²⁺ dynamics in chloroplasts and nongreen plastids. The Cameleon sensors allowed measuring the in vivo Ca²⁺ dynamics in intact organs, like roots and cotyledons/leaves and at single cell level. Importantly, this approach increased the spatial resolution of the analysis to the level of individual chloroplasts, which has not been achievable with aequorin-based approaches for chloroplasts Ca²⁺ analyses (Johnson et al., 1995; Sai and Johnson, 2002; Nomura et al., 2012; Sello et al., 2016).

The 2Bam4-YC3.6 Cameleon probe proved to be preferable for measuring Ca²⁺ dynamics in root plastids as it showed more pronounced ƊR/R₀ responses (amplitudes of transients) to defined eATP stimuli as compared to the 2Bam4-YC4.6 sensor. This is in accordance with the in vitro properties of the YC3.6 sensor, which shows a larger spectroscopic dynamic range than YC4.6 (Nagai et al., 2004; Iwano et al., 2009) as well as a better suitability of its K  _d with the physiologically relevant range of free Ca²⁺ as previously estimated for the plastid stroma (about 50 nm to 1 µm with a consensus of 100–150 nm at baseline; Nomura et al., 2012; Stael et al., 2012; Sello et al., 2016). Attempts of empirical calibration of the sensors in the stroma in vivo proved unsuccessful in our hands, and those of others, due to the impossibility to fully deplete stromal Ca²⁺, which appears to remain protected even when harsh extracellular and cytosolic Ca²⁺ chelation conditions (Fig. 6E; Supplemental Fig. S14 and Supplemental Discussion) was performed. The stromal Ca²⁺ response in nonphotosynthetic plastids correlated in magnitude with cytosolic Ca²⁺ transients triggered by eATP, demonstrating that both compartments responded in a dose-dependent manner and the stromal transient could be abolished by depleting extracellular Ca²⁺ stores. It is tempting to hypothesize that Ca²⁺ is taken up by plastids from the cytosol, hence mirroring the cytosolic transient, while its organellar amplitudes and kinetics are being shaped by import, buffering, and export characteristics. According to this model, nonphotosynthetic plastids may act as buffers and capacitors affecting cytosolic Ca²⁺ transients to allow for specific signatures. Such a situation may arise in loss-of-function mutants of the chloroplast localized Ca²⁺ sensor CAS, which display impaired formation of stimulus-induced cytoplasmic Ca²⁺ signature (Weinl et al., 2008, Nomura et al., 2012).

A key advantage and innovation of the Cameleon toolkit is its applicability at high microscopic resolution to monitor the Ca²⁺ dynamics in chloroplasts of defined tissue areas down to the individual organelle. Studies using aequorin targeted to the stroma have demonstrated that light-to-dark transition induces a gradual and sustained increase in stromal free Ca²⁺ (Sai and Johnson, 2002; Nomura et al., 2012; Sello et al., 2016). We confirmed this observation by analyzing the Cameleon YC3.6 FRET ratio in intact cotyledons averaged over a large number of individual mesophyll chloroplasts (Fig. 4, A and B). However, advancing our analysis to the single chloroplast level allowed refinement of the current picture and identification of at least two response components that take place in guard cell chloroplasts at light-to-low-intensity blue light transitions.

Ca²⁺ spikes in single chloroplasts had not been observed before. The events identified here were not synchronized between different organelles of the same cell. In contrast, the gradual, sustained Ca²⁺ elevation appeared to occur in all chloroplasts in a synchronous manner, despite our finding that this Ca²⁺ elevation has an intraplastidial origin, possibly the thylakoid lumen. The synchronous nature of this response likely reflects that the onset of low-intensity blue light occurred at the same time for all chloroplasts. These results highlight that individual chloroplasts can mount independent Ca²⁺ dynamics. Stromal Ca²⁺ spikes did not mirror cytosolic Ca²⁺ oscillations, which are a characteristic feature of guard cells (Allen et al., 2000; Schroeder et al., 2001; Kudla et al., 2010), as evident by (1) their asynchronous occurrence between individual chloroplasts of a cell, (2) the finding that the extracellular Ca²⁺ chelation could abolish cytosolic oscillations but not stromal spikes, and (3) the fact that application of external Ca²⁺ to permeabilized cells resulted in immediate Ca²⁺ elevation in the cytosol but had practically no discernable effect on chloroplast Ca²⁺ dynamics (Supplemental Fig. S13, A and B).

Our experiments revealed that neither component of chloroplastic Ca²⁺ dynamics (sustained increase nor spikes) appeared to be dependent on apoplastic Ca²⁺ stores. Moreover, while stromal spikes were strongly reduced by the chelation of cytosolic Ca²⁺, as achieved by a combined digitonin and EGTA treatment, the sustained transients in chloroplasts remained unaffected. This indicates that cytosolic Ca²⁺ and/or cellular integrity are the source, or at least the prerequisite for the spikes, while the gradual and sustained increase is solely to depend on an intrachloroplast Ca²⁺ store, such as the thylakoid lumen. Consequently, we conclude that the shaping of stromal Ca²⁺ dynamics occurs by the integration of Ca²⁺ fluxes from two distinct Ca²⁺ sources. Therefore, Ca²⁺ dynamics in the stroma appear to underlie the influence of both the plastid itself, i.e. the former endosymbiont, and rest of the cell, i.e. the former host and extracellular environment of the former endosymbiont. This supports of the concept that Ca²⁺ plays a key role in integrating internal and external stimuli at the level of individual chloroplasts. Interestingly, Ca²⁺ spikes appear under chloroplast-autonomous control, even though the source of the Ca²⁺ causing the spike may be the cytosol.

Spiking in single chloroplasts appears reminiscent of what has been described as “pulses” in membrane potential or pH “flashes” in single mitochondria, which have also been associated with Ca²⁺ (Schwarzländer et al., 2012). Although the electrochemical situations of chloroplast envelope and mitochondrial inner membrane are not directly comparable, the same principles apply and a similar hypothetical working model as put forward for mitochondrial pulses may also explain the stromal Ca²⁺ spikes. Opening of individual Ca²⁺ channels following a stimulus from within the chloroplast itself may allow influx of Ca²⁺ from the cytosol along the negative electrochemical gradient across the chloroplast envelope, which has been demonstrated to be more pronounced in the light than in the dark (Demmig and Gimmler, 1983).

In summary, this work not only establishes a novel toolset to dissect Ca²⁺ dynamics in the plastid stroma in vivo. It moreover demonstrates the complementary power of fluorescent reporter proteins to existing aequorin-based methods by identifying thus far underappreciated complexity of the chloroplast Ca²⁺ dynamics. The latter includes the integration of Ca²⁺ release from chloroplast-autonomous and cytoplasm-originating Ca²⁺ sources to shape stromal Ca²⁺ dynamics with signatures that are specific for the individual organelle. Both the findings and the methods reported pave the way toward the mechanistic dissection of Ca²⁺ signaling in the plastid and the identification of the underlying machinery.

MATERIALS AND METHODS

Plant Material and Growth Conditions

All Arabidopsis (Arabidopsis thaliana) plants were of the ecotype Col-0. Plants were grown on soil under short-day conditions (12 h light/12 h dark, 100 µE m⁻² s⁻¹) at 22°C and 75% relative humidity. Seeds were surface-sterilized by vapor-phase sterilization (Clough and Bent, 1998) and plated on half-strength MS medium (Murashige and Skoog, 1962; Duchefa) supplemented with 0.1% Suc, 0.05% MES, pH 5.8, and 0.8% plant agar (Duchefa). After stratification at 4°C in the dark for 2 d, plates were transferred to the growth chamber under long-day conditions (16 h light/8 h dark, 100 µE m⁻² s⁻¹) at 22°C. The plates were kept vertically and seedlings were used for imaging 7 d after germination. The line expressing the NES-YC3.6 construct was used as reported before (Krebs et al., 2012).

Molecular Cloning and Plasmid Constructs

pGreen 0029 35S:Bam4-YC3.6 was obtained from a modified pGreen0029 plasmid (Hellens et al., 2000), in which the single CaMV35S promoter and the CaMV terminator were inserted by using KpnI and SacI restriction sites, respectively (Bonza et al., 2013). The YC3.6 coding sequence was subcloned from a pCDNA3-YC3.6 construct (Nagai et al., 2004) using the HindIII and EcoRI restriction sites. The chloroplast targeting sequence Bam4 (Fulton et al., 2008) was amplified directly from genomic DNA of Arabidopsis using the following primers: forward 5′-CATGaagcttATGACGGAGACTGGAGTAAT-3′ and reverse 5′-CATGaagcttACGCAACTTAGTGATGAAA-3′ in which HindIII restriction sites (underlined) were inserted at the 5′ and 3′ ends. The sequence was fused to the N terminus of the YC3.6 probe using a HindIII restriction site. To double the Bam4 targeting sequence, the signaling peptide was amplified using the following primers: forward 5′-CATGgggcccATGACGGAGACTGGAGTAAT-3′ and reverse 5′-CATGgggcccACGCAACTTAGTGATGAAA-3′ in which ApaI restriction sites (underlined) were inserted at the 5′ and 3′ ends. The sequence was fused to the N terminus of the Bam4-YC3.6 construct by ApaI restriction. In both the cloning steps, the direction of the insertion was verified by sequencing.

The YC4.6 Cameleon probe was excised from a pRSETB-YC4.6 vector, through digestion with BamHI and EcoRI, and placed in the p35S2 vector (http://www.pgreen.ac.uk/JIT/JIT_fr.htm). The sequence encoding Bam4 was amplified directly from genomic DNA of Arabidopsis using the following primers: forward 5′-CATGaagcttATGACGGAGACTGGAGTAAT-3′ and reverse 5′-CATGggatccCCACGCAACTTAGTGATGAAA-3′ in which a HindIII restriction site (underlined) is present at the 5′ of the forward primer and a BamHI restriction site (underlined) is present at the 5′ of the reverse primer. The sequence was fused to the N terminus of the YC4.6 probe in the p35S2 vector (Hellens et al., 2000) using the HindIII and BamHI restriction sites. The second Bam4 was then amplified using the following primers: forward 5′-CATGaagcttATGACGGAGACTGGAGTAAT-3′ and reverse 5′-CATGaagcttACGCAACTTAGTGATGAAA-3′ in which the HindIII restriction sites (underlined) were inserted at the 5′ and 3′ ends. The sequence was placed between the first Bam4 sequence and the YC4.6 probe by HindIII restriction sites. The entire cassette (35S:2Bam4-YC4.6-Ter) was amplified using the following primers, in which KpnI restriction sites (underlined) are present at the 5′ and 3′ ends of forward 5′-CATGggtaccGATATCGTACCCCTACTCCAAAAAT-3′ and reverse 5′-CATGggtaccGATATCGATCTGGATTTTAGTA-3′. The whole cassette was transferred to the pGreen 0029 vector by KpnI restriction. All DNA amplification by PCR was carried out using Phusion DNA polymerase (Finnzymes). Plasmid amplification was performed using Escherichia coli DH5α cells.

Generation of Transgenic Plants

Plant transformation was carried out using Agrobacterium tumefaciens GV3101 cells by floral dip (Clough and Bent, 1998). Per construct, several independent transgenic lines were selected by antibiotic resistance. The pGreen 0029 35S:2Bam4-YC3.6 construct was introduced both in wild-type Arabidopsis plants and in the rdr6-11 line. The pGreen 0029 35S:2Bam4-YC4.6 and pGPTVII Ubq10:NES-YC3.6 constructs (Krebs et al., 2012) were introduced in the rdr6-11 line. Two or more independent lines of each reporter were used for the experiments.

Chlorophyll Fluorescence Measurements

In vivo chlorophyll fluorescence of Arabidopsis wild-type leaves was measured at room temperature using a Dual PAM-100 fluorometer (Walz) with light at 2,000 µE m⁻² s⁻¹ and actinic light at 166 µE m⁻² s⁻¹. Plants were dark-adapted at room temperature before measurement. The parameter used to measure the efficiency of PSII photochemistry is qP that is reported in plot as 1-qP. qP is a measure of the fraction of open PSII reaction centers and is defined as the coefficients of photochemical fluorescence quenching. Data are presented as averages ± sd of three independent leaves/experiments from independent plants.

CLSM

CLSM analyses were performed using a Leica SP5 imaging system and a Zeiss LSM780 (Carl Zeiss Microscopy). On the Leica SP5, cpVenus and chlorophyll were excited by the 514-nm line of the argon laser, and the emission was collected at 525 to 540 nm and 650 to 750 nm, respectively. Images were acquired by a 25× water immersion objective with different digital zoom. On the Zeiss LSM780, a 40× (C-Apochromat, 1.20 N.A., water immersion) or 63× lens (Plan-Apochromat, 1.40 N.A., oil immersion) was used. CpVenus and chlorophyll were excited at 514 nm, and fluorescence was measured at 525 to 550 nm and 647 to 745 nm, respectively, with the pinhole set to 1 airy unit. Ca²⁺ imaging of intact 7- to 8-d-old seedlings was performed as recently described (Wagner et al., 2015a, 2015b). YC3.6 was excited at 458 nm, and emission of FRET pair proteins ECFP and cpVenus was collected at 465 to 500 nm and 525 to 560 nm, respectively. For the evaluation of possible spectroscopic effects of chlorophyll, the chlorophyll autofluorescence was excited with 458 nm and detected with the band-pass filter of 647 to 745 nm. Images were analyzed using ImageJ software.

Fluorescence Microscopy

Roots, cotyledons, and guard cells of the Cameleon reporter lines were analyzed in vivo using an inverted fluorescence microscope (Nikon Ti-E). The objectives used were a 20× (CFI Plan APO 20× VC, N.A. 0.75) dry objectives and a 60× (CFI Plan APO Lambda 60× N.A. 1.4) oil immersion. Excitation light was produced by a fluorescent lamp Prior Lumen 200 PRO (Prior Scientific) at 440 nm (436/20 nm) set to 20%. Images were collected with a Hamamatsu ORCA-D2 Dual CCD camera. For Cameleon analysis, the FRET CFP/YFP optical block A11400-03 (emission 1 483/32 nm for CFP and emission 2 542/27 nm for the FRET) with a dichroic 510-nm mirror (Hamamatsu Photonics) was used for the simultaneous CFP and FRET acquisitions (cpVenus for YC3.6 and YC4.6). Exposure time was from 100 to 400 ms with 2 × 2 CCD binning for cytosolic Cameleon (wild-type NES-YC3.6 and rdr6 NES-YC3.6) and a 4 × 4 CCD binning for chloroplastidial/plastidial Cameleon (rdr6 2Bam4-YC3.6 and rdr6 2Bam4-YC4.6). Images where acquired in 5-s intervals (20-s intervals after the first 10 min of recording in the experiments of Fig. 5, C–E). Filters and dichroic mirrors were purchased from Chroma Technology. The NIS-Element (Nikon) was used as a platform to control microscope, illuminator, camera, and postacquisition analyses.

Root and Cotyledon Seedling Imaging

Seven-day-old seedlings were used for root and cotyledon imaging. Seedlings were kept in the growth chamber until the experiment. For root experiments, the seedlings were gently removed from the plate according to Behera and Kudla (2013), placed in the dedicated chambers, and overlaid with cotton wool soaked in imaging solution (5 mm KCl, 10 mm MES, and 10 mm CaCl₂, pH 5.8 adjusted with Tris-base). The root was continuously perfused with imaging solution while the shoot was not submerged. The eATP treatment, by supplementing the imaging solution with 10, 100, or 500 µm ATP, was administered for 3 min under running perfusion. For cotyledon experiments, the same procedure was adopted, but the shoot was then also submerged and perfused. Before the measurements were started, the bright-field light of the microscope was kept on at 100 µE m⁻² s⁻¹ for 5 min. Depending on the objective the fluorescent light flux was 100 µE m⁻² s⁻¹ (20×) or 0.26 µE m⁻² s⁻¹ (60×).

Guard Cell Imaging

A small leaf piece (approximately 10 mm²) of a 4-week-old Arabidopsis plant was glued to the cover slide using medical adhesive (Hollister). Upper cell layers were gently removed with a razor blade. The epidermal strips obtained were incubated in the guard cell solution (5 mm KCl, 10 mm MES, and 50 μm CaCl₂, pH 6.15 adjusted with Tris-base) and placed under bright-field light (100 µE m⁻² s⁻¹) in the growth chamber for 5 to 10 min, before starting the measurements or treatments.

Quantitative Imaging Analysis

Fluorescence intensity was determined over regions of interest that corresponded to the root tip meristematic zone, entire field of view in cotyledons, and single guard cells or single chloroplast. FRET and CFP emissions of the analyzed regions of interest were used for the ratio (R) calculation (FRET/CFP) and, where suitable, normalized to the initial ratio (R₀) and plotted versus time (ƊR/R₀). Background subtraction was performed independently for both channels before calculating the ratio, with exception of the cotyledon experiments, where no image area allowed determination of the background. For the transformation of FRET ratio data to absolute free Ca²⁺ concentrations, the following assumptions were made: (1) cytosolic baseline of free Ca²⁺ of 108 nm in Col-0 (Logan and Knight, 2003), (2) in vitro behavior of the sensor (K  _d 250 nm, Hill coefficient 1.7, dynamic range 5.32 at pH 8.0 [plastid stroma] and 5.43 at pH 7.4 [cytosol]), and (3) comparable physicochemical sensor behavior in the stroma and in the cytosol as recently described in Wagner et al. (2015b) following Palmer and Tsien (2006).

Ca²⁺ Chelation and Guard Cell Permeabilization

To sequester extracellular Ca²⁺ stores, leaf epidermal strips prepared for guard cell imaging were placed in the EGTA solution (guard cell solution supplemented with 1 mm EGTA) and left in the growth chamber in the light for 5 min before starting the measurement. To permeabilize cells and sequester Ca²⁺, leaf epidermal strips were placed in permeabilization solution (100 mm K-gluconate, 1 mm MgCl₂, 10 mm HEPES, 5 mm EGTA, and 500 µM digitonin, pH 7.5 adjusted with Tris-base) for 4 min. The samples were then placed in maintaining solution (100 mm K-gluconate, 1 mm MgCl₂, 10 mm HEPES, and 1 mm EGTA, pH 7.5 adjusted with Tris-base) and left in the growth chamber under bright field light (100 µE m⁻² s⁻¹) for 5 min before starting the measurement.

Accession Numbers

Sequence data from this article can be found in the GenBank/EMBL data libraries under the following accession numbers: Bam4, AT5G55700; RDR6, AT3G49500; YC3.6, AB178712; and YC4.6, AB178713.

Supplemental Data

The following supplemental materials are available.

• Supplemental Figure S1. Subcellular localization of Bam4-YC3.6 and 2Bam4-YC3.6 sensor fluorescence in tobacco leaf epidermal cells.

• Supplemental Figure S2. Expression of 2Bam4-YC3.6 probe in Arabidopsis transgenics with wild-type background.

• Supplemental Figure S3. Subcellular localization of 2Bam4-YC3.6.

• Supplemental Figure S4. Expression of 2Bam4-YC4.6 probe in Arabidopsis transgenics with rdr6 background.

• Supplemental Figure S5. Expression of NES-YC3.6 probe in Arabidopsis transgenics with rdr6 background.

• Supplemental Figure S6. Single wavelength emission of cpVenus (yellow trace) and CFP (light blue trace) of the 2Bam4-YC3.6 and 2Bam4-YC4.6 probes used for the ratio calculations shown in Figure 2.

• Supplemental Figure S7. Plastidial and cytosolic Ca²⁺ transients in response to different concentrations of eATP.

• Supplemental Figure S8. The effect of light-to-low-intensity blue light illumination transitions on plastidial Ca²⁺ levels.

• Supplemental Figure S9. Pulse amplitude modulation of Arabidopsis wild-type leaves.

• Supplemental Figure S10. Chloroplast “spiking” in Arabidopsis mesophyll cells and stromal Ca²⁺.

• Supplemental Figure S11. The 2Bam4-YC3.6 reports Ca²⁺ levels only in chloroplasts.

• Supplemental Figure S12. 2Bam4-YC3.6 reports a low-intensity blue light illumination-induced accumulation of stromal Ca²⁺.

• Supplemental Figure S13. Cytosolic Ca²⁺ dynamics of guard cells measured in wild-type NES-YC3.6 and rdr6 NES-YC3.6 plants in response to light-to-low-intensity blue light illumination transitions.

• Supplemental Figure S14. In vivo time-lapse digitonin-dependent permeabilization of rdr6 NES-YC3.6 and rdr6 2Bam4-YC3.6 guard cells.

• Supplemental Movie S1. Ratiometric false-color movie from a representative time series of an rdr6 cotyledon expressing the 2Bam4-YC3.6 probe in response to a light-to-low-intensity blue light illumination transition.

• Supplemental Movie S2. Ratiometric false-color movie from a representative time series of an rdr6 guard cell expressing the 2Bam4-YC3.6 sensor in response to a light-to-low-intensity blue light illumination transition.

• Supplemental Discussion.

ACKNOWLEDGMENTS

The rdr6-11 mutant line was kindly provided Roberto De Michele (IBBR, CNR, Palermo, Italy), the pRSETB-YC4.6 vector by Atsushi Miyawaki (RIKEN Brain Science Institute, Japan), and the transgenic NES-YC3.6 Arabidopsis line by Karin Schumacher (University of Heidelberg, Germany). We thank Tomas Morosinotto (University of Padova, Italy) for help with the PAM imaging experiments and Carol Priestley for feedback on the manuscript.

Glossary

 
• CLSM

  confocal laser scanning microscopy

 
• eATP

  extracellular ATP

 
• ER

  endoplasmic reticulum

 
• PAM

  pulse amplitude modulation

Author notes

1

This work was supported by Ministero dell’Istruzione, dell’Università e della Ricerca Fondo per gli Investimenti della Ricerca di Base (FIRB) 2010 RBFR10S1LJ_001 grant to A.C., by Regione Lombardia “Filagro,” “Progetto di Ricerca di Ateneo” (CPDA122838) to M.Z., by a binational DAAD/VIGONI grant to M.Z. and J.K., and by a grant from the Deutsche Forschungsgemeinschaft (DFG; FOR964) to J.K. M.S. thanks the DFG for funding through the Emmy-Noether program (SCHW1719/ 1-1), the Research Training Group GRK2064, and a grant (SCHW1719/5-1) as part of the package PAK918.

*

Address correspondence to .

The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (www.plantphysiol.org) is: Alex Costa ().

A.C., M.Z., and J.K., designed research; A.C., G.L., S.Wa, F.G.D., S.B., and S.We. performed the experiments; S.Wa. performed the confocal microscope experiments with the Zeiss 780; A.C., M.Z., J.K., and M.S. analyzed data, supervised the research, and wrote the article.

www.plantphysiol.org/cgi/doi/10.1104/pp.16.00652