Generation of Red-Shifted Cameleons for Imaging Ca²⁺ Dynamics of the Endoplasmic Reticulum

Abstract

Cameleons are sophisticated genetically encoded fluorescent probes that allow quantifying cellular Ca²⁺ signals. The probes are based on Förster resonance energy transfer (FRET) between terminally located fluorescent proteins (FPs), which move together upon binding of Ca²⁺ to the central calmodulin myosin light chain kinase M13 domain. Most of the available cameleons consist of cyan and yellow FPs (CFP and YFP) as the FRET pair. However, red-shifted versions with green and orange or red FPs (GFP, OFP, RFP) have some advantages such as less phototoxicity and minimal spectral overlay with autofluorescence of cells and fura-2, a prominent chemical Ca²⁺ indicator. While GFP/OFP- or GFP/RFP-based cameleons have been successfully used to study cytosolic and mitochondrial Ca²⁺ signals, red-shifted cameleons to visualize Ca²⁺ dynamics of the endoplasmic reticulum (ER) have not been developed so far. In this study, we generated and tested several ER targeted red-shifted cameleons. Our results show that GFP/OFP-based cameleons due to miss-targeting and their high Ca²⁺ binding affinity are inappropriate to record ER Ca²⁺ signals. However, ER targeted GFP/RFP-based probes were suitable to sense ER Ca²⁺ in a reliable manner. With this study we increased the palette of cameleons for visualizing Ca²⁺ dynamics within the main intracellular Ca²⁺ store.

1. Introduction

The endoplasmic reticulum (ER) exhibits a 1000-fold higher concentration of Ca²⁺ than that in the cytosol or mitochondria by storing more than 90% of intracellular Ca²⁺ in less than 10% of total cell volume [1,2]. The release of Ca²⁺ from this internal Ca²⁺ store is important for multiple signaling events [3] that regulate ATP production [4], insulin secretion [5], muscle contraction [6,7] or gene regulation [8,9]. Based on the importance of the ER Ca²⁺ content for cell signaling, processes that refill the ER with Ca²⁺ are necessary. Particularly, in non-excitable cells, the release of Ca²⁺ from the ER induces the so-called store operated Ca²⁺ entry (SOCE). SOCE is based on the interaction of the stromal interacting molecule 1 (STIM1) with the Ca²⁺ channel pore-forming unit ORAI1 in the plasma membrane [10]. The fundamental role of SOCE is to restore Ca²⁺ of the ER during and after cell stimulation. Interestingly, in endothelial cells mitochondria are important to direct entering Ca²⁺ to the ER during the process of ER Ca²⁺ refilling [11,12,13]. In addition to the importance of ER Ca²⁺ for the generation of global and local Ca²⁺ signals, the ER Ca²⁺ content considerably impacts the protein folding machineries within this organelle [14]. A loss of Ca²⁺ within the ER is associated with the induction of the unfolding protein response (UPR), which can also lead to cell death [15]. Recently, it was shown that the energy supply of the ER is tightly coupled to the organelles’ Ca²⁺ homeostasis. In order to investigate the dynamic storage of Ca²⁺ by the ER and to correlate changes of the ER Ca²⁺ content with other ER-related signaling events under physiological as well as pathophysiological conditions, the demand on reliable ER Ca²⁺ sensors is obvious.

Although some chemical indicators have been successfully established for assessing the ER Ca²⁺ homeostasis [13,16], these requirements were better realized by sophisticated genetically encoded Ca²⁺ indicators (GECIs). In general, GECIs can be categorized into two classes: the FRET based cameleons that consists of two fluorescent proteins (FP) and the single FP-based GCaMP-like sensors [17,18]. Both types take different advantages, e.g., the cameleons benefit from ratiometric signal changes, photo- and pH-stability [19,20], whereas single FP-based indicators have larger signal changes, better dynamic kinetics and they are more variable in their fluorescent hues [21,22,23,24,25]. Cameleon-type indicators are constructed of a genetic fusion of calmodulin (CaM) and a small Ca²⁺-CaM binding peptide (M13), which is flanked by the donor and acceptor fluorescent proteins. The Ca²⁺-dependent interaction of CaM with M13 induces a conformational change that narrows the distance of the two FPs and, thus, results in an increase of the FRET ratio signal [26] In contrast, in GCaMP-type indicators a circular permutated FP (cpFP) is flanked by M13 at its N-terminus and CaM at its C-terminus. Herein, the interaction of the Ca²⁺ dependent motifs changes the environment of the cpFP’s chromophore inducing an intensiometric increase of the fluorescence [27]. Moreover, the cameleons are more accessible to genetically shift the balance of the dissociation constant (K_d) [28,29], but, due to the presence of two different FPs, the possibilities for multi-color imaging are limited [30,31]. For measuring Ca²⁺ signals of the ER, both classes of GECIs have been conjugated with a calreticulin signal [32] and the classical ER retention sequence [33] which target the probes into the ER lumen.

However, so far most of GECIs were used to detect Ca²⁺ in the cytosol or mitochondria. By mutating the original CaM-M13 domain in order to reduce the Ca²⁺ binding affinity the first ER Ca²⁺ cameleons were developed. So far a cameleon named D1ER represents the state-of-the-art probe for sensing [Ca²⁺]_ER [34] This probe consists of an enhanced CFP (ECFP) as FRET donor and the yellow FP Citrine as FRET acceptor flanking the Ca²⁺ sensitive motif D1 (design 1). In D1 the CaM and M13 domains were redesigned to optimize the Ca²⁺ affinity for quantifying ER Ca²⁺ dynamics.

In this study, we engineered numerous red-shifted ER-targeted cameleons based on the optimized D1 motif. Such bathochromic probes can be better used for simultaneous measurements of [Ca²⁺]_ER and [Ca²⁺]_Cyto by combining these novel GECIs with fura-2. Our results show that despite an ER targeting and ER retention sequences, cameleons containing cpEGFP on the C-terminus show poor ER localization. Moreover, the GFP/OFP-based ER targeted cameleons gained Ca²⁺ sensitivity, which make these probes impractical for sensing ER Ca²⁺ signals. Finally, we found suitable GFP/RFP FRET pairs for the development of functional ER targeted cameleons, which are suitable for dual recordings of [Ca²⁺]_ER and [Ca²⁺]_Cyto in single individual cells.

2. Experimental Section

2.1. Materials

For culturing cells, RPMI-1640, fetal calf serum (FCS), penicillin, streptomycin, sodium pyruvate, vitamins, essential amino acids and all plastic ware were obtained from PAA laboratories (Pasching, Austria). Dulbecco’s modified eagle’s medium (DMEM), histamine dihydrochloride (histamine), 2,5-di-t-butyl-1,4-benzohydroquinone (BHQ) and digitonin were purchased from Sigma-Aldrich (Vienna, Austria). Ionomycin (free acid) was delivered from abcamBiochemicals (Cambridge, UK), fura-2/AM (Fura2) from Teflabs (Texas Fluorescence Labs Inc., Austin, TX, USA) and T₄ DNA Ligase from Thermo Scientific (Braunschweig, Germany). TagRFP and Camui-CR plasmids were purchased from Addgene. GoTaq^® Hot Start Polymerase, restriction enzymes, Wizard^® SV gel and PCR Clean-up System, PureYield^TM Plasmid Maxiprep System and Transfast^TM transfection reagent were obtained from Promega (Mannheim, Germany). All other materials and chemicals were from Roth (Karlsruhe, Germany).

2.2. Engineering of Sensors

For plasmid construction of the various genetically encoded Ca²⁺ indicators (GECIs) a pcDNA3.1(-) vector containing the coding sequences (cds) of the recently published cameleons D1GO-Cam or mtD1GO-Cam were used as templates. Substitution of the cds from the different fluorescent proteins (FP) was done by PCR amplification, restriction digestion and ligation to the N- or C-terminal end of the sensor sequence. Targeting of cameleons to the ER was achieved by primer annealing and subsequent ligation of the ER targeting sequence from calreticulin to the N-terminal end of the GECIs. For proper ER retention, the coding sequence lysine-aspartic acid-glutamic acid-leucine (KDEL) was added to the reverse primers of the C-terminal inserted fluorescent protein sequence. A list of all primers and restriction enzymes used is given in Table S1.

2.3. Cell Culture and Transfection

Human embryonic kidney 293 stably expressing ryanodine receptor 2 (HEK E2) [35] and HeLa cells (HeLa) [36] were cultured in DMEM supplemented with 10% FCS, 100 U/mL penicillin, 100 µg/mL streptomycin. INS-1 832/13 cells (INS-1) [5] kindly obtained from C. B. Newgard (Duke University School of Medicine, Durham, NC, USA) were cultured with RPMI-1640 medium containing 10% FCS, 11.1 mM D-glucose, 1 mM sodium pyruvate, 5 µM mercaptoethanol, 100 U/mL penicillin and 100 µg/mL streptomycin. The 3 different cell lines were maintained in a cell incubator at 37 °C in a humidified atmosphere with 5% CO₂ and 95% air. For transfection and experiments cells were cultured on glass cover slips (Ø = 30 mm). Transfection was performed at 50%–80% cell confluence by adding a transfection mixture of DMEM supplemented with the respective sensor plasmid DNA and Transfast^TM transfection reagent. After 12–18 h incubation, transfection mixture was replaced by normal culture medium. Imaging experiments were performed 40–50 h after transfection.

2.4. Fura-2 Loading and Experimental Buffers for Ca²⁺ Recordings

For dual recordings of [Ca²⁺]_Cyto and [Ca²⁺]_ER cells expressing one of the red-shifted cameleons were loaded with 3 µM fura-2/AM at room temperature for 30–40 min in a buffer composed of (in mM): 2 CaCl₂ 135 NaCl, 5 KCl, 1 MgCl₂, 1 HEPES, 2.6 NaHCO₃, 0.44 KH₂PO₄, 0.34 Na₂HPO₄, 10 D-glucose, 0.1% vitamins, 0.2% essential amino acids and 1% penicillin/streptomycin pH 7.4. Fura-2 loaded cells as wells as non-loaded cells were washed and stored in this buffer prior to imaging experiments. For Ca²⁺ measurements, cells were perfused in a HEPES buffered solution containing (in mM): 140 NaCl, 5 KCl, 1 MgCl₂, 1 HEPES and 10 D-Glucose pH 7.4 (0 mM Ca²⁺). Additionally, either 1 mM EGTA (EGTA), 1 mM CaCl₂ (1 mM Ca²⁺) or 2 mM CaCl₂ (2 mM Ca²⁺) were added to 0 mM Ca²⁺ and pH adjusted to 7.4.

2.5. Spectral Scans and K_d Determination

HeLa cells transfected with ER targeted cameleons were permeabilized with 5 µM digitonin and 10 µM ionomycin (digitonin/ionomycin) in 2 mM Ca²⁺ buffer for 3 min. For spectral scans, EGTA buffer supplemented with digitonin/ionomycin was added for another 3 min to obtain maximum depletion of Ca²⁺ from the ER. Individual red-shifted ER sensors were alternately exposed at 400 to 490 nm in 10 nm steps for 400 ms each through the whole experimental time course. Emitting light of FRET donor and FRET acceptor was recorded at 510 and 560 nm, respectively, as previously described [37] Signal to noise ratios were then calculated at all excitation wavelengths for each sensor. Dissociation constant (K_d) of [Ca²⁺] from D1ERCmR2 was determined in situ. The ER of permeabilized HeLa cells was depleted using a 3 mM EGTA buffer. Free Ca²⁺ concentrations were calculated using the CaBuff software (G. Droogmans, Fysiologie, Leuven) and appropriate amount of Ca²⁺ were added to the 3 mM EGTA buffer to obtain buffer solutions of 1, 3, 10, 30, 100, 300, 1000, 2000 and 10,000 µM free Ca²⁺.

2.6. Ca²⁺ Recordings and Data Acquisition

Ca²⁺ imaging was done using the TiLL iMIC (Till Photonics, Graefelfing, Germany) digital wide field imaging system, as described previously [12,31,38,39]. Dual Ca²⁺ recording of fura-2 and red shifted cameleons was performed by alternated excitations at 340, 380 and 480 nm and emissions were captured at 510 and 560 nm. D1ER was excited at 430 nm and emission was collected using the dichrotome dual emission filterset (dichroic 535dcxr, CFP emitter 482/18 nm and YFP emitter 535/3 nm). Data acquisition and control of the digital fluorescence microscope was done using the live acquisition software version 2.0.0.12 (Till Photonics). Results of FRET measurements are either shown as the ratio of (F₅₃₅/F₄₈₀)/R₀ for D1ER or (F₅₆₀/F₅₁₀)/R₀ for red-shifted cameleons.

2.7. Confocal Imaging

Images of subcellular structures for colocalization were taken from cells coexpressing D1ER and either D1ERGO-Cam1 or D1ERGO-Cam2. Fluorescence of D1ERCmR2 expressing cells was either imaged alone or together after fura-2/AM loading. All images were recorded with an array confocal laser scanning microscope (ACLSM) built on a fully automatic inverse microscope (Axio Observer.Z1, Zeiss, Göttingen, Germany) equipped with VoxCell Scan^® (VisiTech, Visitron Systems) using a 100 × objective (Plan-Fluor 100 × /1.45 oil, Zeiss), as described previously [12,37,40]. Excitation was done using laser light of diode lasers (Visitron Systems): Fura-2 was excited at 405 nm (120 mW), CFP of D1ER was excited at 445 nm (50 mW), Clover of D1ERCmR2 was excited at 473 nm (50 mW), mKO and mRuby2 of D1ERGO-Cam1, D1ERGO-Cam2 and D1ERCmR2 were excited at 515 nm (50 mW). Emitted light was acquired with emission filters ET460/50m for fura-2 (DAPI filter), ET480/40m for CFP, ET525/50m for Clover and E570LPv2 for mKO and mRuby2 (Chroma Technologies, Corporation, VT, USA). A Photometrics CCD camera (CoolSnap HQ2) was used to capture all images. Quantitative ER colocalization computations were performed with the integrated morphometric analysis plug-in of MetaMorph 7.7.0.0 software (Visitron).

2.8. Statistics

Statistical relevant data are shown as means ± SEM, where n represents the number of cells. Analyses were done using unpaired Student’s t test and evaluation of significance was considered to be p < 0.05.

3. Results and Discussion

3.1. GFP/OFP FRET-Based Cameleons are Unsuitable to Measure Ca²⁺ Dynamics of the ER

Analogous to recently developed red-shifted cameleons for imaging Ca²⁺ signals either within the cytosol (D1GO-Cam) or mitochondria (mtD1GO-Cam) [31] we constructed the ER targeted Ca²⁺ probe D1ERGO-Cam1 and D1ERGO-Cam2 (Figure 1A). For ER targeting the respective red-shifted cameleon was fused with the targeting sequence of calreticulin on the N-terminus and a KDEL retention sequence was added at the C-terminus, respectively. However, ER localization of D1ERGO-Cam1 was poor in the rat pancreatic beta cell line, INS-1 (Figure 1B). Colocalization analysis revealed that only 23.46% ± 1.89% (n = 10) of the D1ERGO-Cam1 fluorescence merged with the D1ER signal in this cell type. The localization of D1ERGO-Cam1 in non-ER compartments might be due to the circularly permutated enhanced green fluorescent protein (cpEGFP) on the C-terminus, as it has been reported that circularly permutated Venus is not localized well in the ER [17]. Hence, we speculate that circularly permutated FPs mask the ER retention sequence, so that these constructs are exported from the organelle. Indeed, the exchange of cpEGFP by EGFP in D1ERGO-Cam2 (Figure 1A) considerably improved the ER targeting of the red-shifted cameleon (Figure 1B). D1ERGO-Cam2 co-localized by 83.34% ± 2.33% (n = 10) with D1ER in INS-1 cells. However, both red-shifted cameleons D1ERGO-Cam1 and D1ERGO-Cam2 containing cpEGFP or EGFP as the FRET donor and the monomer Kusabira Orange (mKO) as the FRET acceptor hardly sensed Ca²⁺ dynamics within the ER (Figure 1C). The poor performance of these red-shifted probes to report changes in ER Ca²⁺ is most likely due to an increased sensitivity of the cameleons to bind Ca²⁺. In a recent study we demonstrated that despite the reported dissociation constant (K_d) of approximately 200 µM of the D1 domain in D1ER, exchanging the CFP/YFP FRET pair by GFP/OFP and removing the ER targeting sequence resulted in a shift of the K_d to 1 µM [31] Accordingly, these red-shifted D1 cameleons work well in the cytosol and mitochondria. In this study, we targeted the GFP/OFP-based D1 cameleons in the ER lumen in order to test if the localization within this organelle impacts the Ca²⁺ binding affinity of the probe. Our results, however, confirmed that independent of the subcellular localization GFP/OFP-based D1 cameleons have a significantly increased Ca²⁺ sensitivity. Next we constructed numerous other red-shifted ER targeted cameleons that might have lower sensitivities and, hence, suitable for visualizing dynamic changes of [Ca²⁺]_ER.

Figure 1. Evaluation of GFP/OFP cameleons upon ER targeting: (A) schematic composition of D1ERGO-Cam1 and D1ERGO-Cam2; and (B) representative confocal images for colocalization analysis. INS-1 cells were co-transfected with D1ER and either D1ERGO-Cam1 or D1ERGO-Cam2, respectively. Images of GFP/OFP-based cameleons were taken from the mKO emission at 570 nm (left panels); D1ER was monitored in its CFP emission at 480 nm (middle panels); and an overlay of images (right panels); (C) Average curves of FRET measurements in HeLa cells expressing either D1ER (cyan curve, n = 14), D1ERGO-Cam1 (dark grey curve, n = 10) or D1ERGO-Cam2 (green curve, n = 7) upon treatment with 100 µM histamine and 15 µM BHQ in the absence of extracellular Ca²⁺. SOCE was accomplished by the subsequent addition of 2 mM Ca²⁺.

Figure 1. Evaluation of GFP/OFP cameleons upon ER targeting: (A) schematic composition of D1ERGO-Cam1 and D1ERGO-Cam2; and (B) representative confocal images for colocalization analysis. INS-1 cells were co-transfected with D1ER and either D1ERGO-Cam1 or D1ERGO-Cam2, respectively. Images of GFP/OFP-based cameleons were taken from the mKO emission at 570 nm (left panels); D1ER was monitored in its CFP emission at 480 nm (middle panels); and an overlay of images (right panels); (C) Average curves of FRET measurements in HeLa cells expressing either D1ER (cyan curve, n = 14), D1ERGO-Cam1 (dark grey curve, n = 10) or D1ERGO-Cam2 (green curve, n = 7) upon treatment with 100 µM histamine and 15 µM BHQ in the absence of extracellular Ca²⁺. SOCE was accomplished by the subsequent addition of 2 mM Ca²⁺.

3.2. Development of Optimized Red-Shifted Cameleons for Imaging ER Ca²⁺ Dynamics Using GFP/RFP FRET Pairs

Under consideration of the fluorescent properties of various FPs, we selected several green FRET-donor and red FRET-acceptor proteins for the generation of novel functional red-shifted ER targeted cameleons (Table S2). For this purpose, we also included the green Clover and the red mRuby FP that had been recently used as an optimized FRET pair for the development of red-shifted genetically encoded probes for kinase activities and the membrane potential [41] Moreover the N- or C-terminal orientation of the FRET-donor and -acceptor FPs within ER targeted cameleons was considered as a relevant determinant of the sensors’ properties (ER localization and Ca²⁺ sensitivity). In total, we generated and tested 10 different ER targeted GFP/RFP-based cameleons that differ either in the FP-FRET pair or the orientation of FPs within the construct (C- or N-terminal location, Table 1). To test the properties of the novel red-shifted ER targeted cameleons, HeLa cells expressing the probes were imaged. During imaging at different excitation wavelengths ranging from 400 to 490 nm, ER Ca²⁺ levels were manipulated by cell treatment with digitonin/ionomycin in the presence and absence of Ca²⁺ in the experimental buffer (Figure 2A,B). Out of these spectral scans, the signal to noise ratio (SNR) for each sensor at each excitation wavelength was calculated to evaluate the best working red-shifted cameleon in the ER under these conditions (Table 2). All the GFP/RFP-based sensors were suitable to detect changes of [Ca²⁺]_ER (Figure 2). D1ERTG (Table 1 and Figure 2A), which was built of the bright tandem dimer Tomato as the FRET acceptor and EGFP as the FRET donor at the C-terminal end, appeared with the highest basal FRET ratio (Figure 2A), while the SNR was rather low (Table 2). Interestingly, the orientation of the FPs within the EGFP/tagRFP-consisting cameleons influenced the SNR significantly. If the tagRFP FRET acceptor was on the N-terminus of the ER targeted cameleon, better signals could be obtained (Figure 2A and Table 2). However, the cameleon with Clover on its N-terminal end and mRuby2 on its C-terminal end, named D1ERCmR2, showed the best performance with the highest SNR (Table 1 and Table 2; Figure 2B). This finding is in line with the original report about the usage of the Clover/mRuby FRET pair in genetically encoded probes [41] These are reporters of kinase activity (Camuiα-CR, AKAR-CR), small GTPase activity (Raichu-CR) and transmembrane voltage (VSFP-CR) [41]. Within all these indicators, the standard FRET pair CFP/YFP has been replaced by Clover as donor at the N-terminal end and by mRuby2 as FRET acceptor at the C-terminal end. Moreover, Clover and mRuby2 were characterized as the brightest green and red FPs exhibiting the highest Förster radius of any ratiometric FRET pair to date. We also tested if the orientation of Clover and mRuby in ER targeted cameleons impact the performance of the probe. In this case, the exchange of the FPs resulting in the construct D1ERmR2C (with mRuby2 on the N-terminus and Clover on the C-terminus) dramatically worsened the respective signal (Figure 2B; compare red curve with blue curve and respective SNR values in Table 2). Similar results were also found in intact HeLa cells that were treated for several minutes with histamine and the reversible sarcoplasmic/endoplasmic reticulum calcium ATPase (SERCA) inhibitor 2,5-di-tert-butylhydroquinone (BHQ) to transiently deplete the ER Ca²⁺ content (Figure 3). In line with the data obtained from the spectral scans, D1ERCmR2 showed a significantly higher delta FRET ratio signal in comparison with all other ER targeted probes in intact cells under these conditions (Figure 3). These findings indicate that in GFP/RFP-based cameleons that contain D1 as the Ca²⁺ sensing domain keep the high K_d of approximately 200 µM and are, hence, suitable to monitor Ca²⁺ dynamics of the ER. However, the origin and orientation of FPs forming the FRET pair within ER targeted cameleons have a huge impact on FRET ratio signals in response to ER Ca²⁺ mobilization.

Table 1. List of GFP/RFP FRET-based cameleons targeted to the ER.

Sensor Name	Schematic Overview
D1ERRG-Cam1
D1ERRG-Cam2
D1ERGR
D1ERTG
D1ERRC
D1ERCR
D1ERmR2G
D1ERGmR2
D1ERmR2C
D1ERCmR2

Table 1. List of GFP/RFP FRET-based cameleons targeted to the ER.

Sensor Name	Schematic Overview
D1ERRG-Cam1
D1ERRG-Cam2
D1ERGR
D1ERTG
D1ERRC
D1ERCR
D1ERmR2G
D1ERGmR2
D1ERmR2C
D1ERCmR2

Figure 2. Functional evaluation of red-shifted cameleons in the ER. Permeabilized HeLa cells transfected with individual red-shifted sensors targeted to the ER were recorded in an extracellular Ca²⁺-free solution and in a 2 mM Ca²⁺ containing environment, respectively. Representative curves showing FRET signals of each indicator excited at 420, 450, 460, 470, 480 or 490 nm and emissions were collected from either (A) D1ERGO-Cam1 (dark grey curve), D1ERGO-Cam2 (green curve), D1ERRG-Cam1 (pink curve), D1ERRG-Cam2 (purple curve), D1ERGR (light grey curve), D1ERTG (violet curve) or (B) D1ERRC (light green curve), D1ERCR (grey curve), D1ERGmR2 (black curve), D1ERmR2G (dark red curve), D1ERmR2C (blue curve) or D1ERCmR2 (red curve).

Figure 2. Functional evaluation of red-shifted cameleons in the ER. Permeabilized HeLa cells transfected with individual red-shifted sensors targeted to the ER were recorded in an extracellular Ca²⁺-free solution and in a 2 mM Ca²⁺ containing environment, respectively. Representative curves showing FRET signals of each indicator excited at 420, 450, 460, 470, 480 or 490 nm and emissions were collected from either (A) D1ERGO-Cam1 (dark grey curve), D1ERGO-Cam2 (green curve), D1ERRG-Cam1 (pink curve), D1ERRG-Cam2 (purple curve), D1ERGR (light grey curve), D1ERTG (violet curve) or (B) D1ERRC (light green curve), D1ERCR (grey curve), D1ERGmR2 (black curve), D1ERmR2G (dark red curve), D1ERmR2C (blue curve) or D1ERCmR2 (red curve).

Table 2. Signal to noise ratios of various ER targeted red-shifted cameleons at stepwise increased excitation wavelengths. Optimum excitations for individual sensors resulting in highest SNR are highlighted.

Sensor	400	410	420	430	440	450	460	470	480	490
D1ERGO-Cam1	0.39	1.90	1.91	4.00	4.41	5.04	6.88	11.26	12.14	6.03
D1ERGO-Cam2	2.12	2.87	3.43	5.72	6.43	10.07	18.49	28.21	30.96	32.32
D1ERRG-Cam1	19.51	25.93	32.90	28.92	41.51	29.91	47.34	57.18	56.88	48.52
D1ERRG-Cam2	29.96	43.93	57.30	84.84	72.16	84.43	70.16	77.38	74.34	66.53
D1ERGR	4.90	6.99	7.53	12.17	16.09	27.44	40.84	55.94	56.59	41.48
D1ERTG	15.08	19.64	37.10	25.63	29.60	38.43	44.83	42.29	41.86	37.52
D1ERRC	2.47	3.57	5.06	8.33	14.23	23.26	47.45	81.02	92.87	92.57
D1ERCR	6.71	5.79	6.49	12.40	19.04	42.11	77.67	142.01	133.05	123.24
D1ERmR2G	9.84	13.56	17.92	22.72	25.01	50.53	82.87	107.33	115.81	107.91
D1ERGmR2	9.36	10.94	15.66	18.13	26.91	44.56	58.36	70.51	63.02	60.20
D1ERmR2C	1.99	2.26	2.74	3.68	5.21	11.83	18.93	30.48	31.09	40.35
D1ERCmR2	16.70	15.85	17.65	33.00	51.42	85.53	151.17	246.90	263.13	246.64

Table 2. Signal to noise ratios of various ER targeted red-shifted cameleons at stepwise increased excitation wavelengths. Optimum excitations for individual sensors resulting in highest SNR are highlighted.

Sensor	400	410	420	430	440	450	460	470	480	490
D1ERGO-Cam1	0.39	1.90	1.91	4.00	4.41	5.04	6.88	11.26	12.14	6.03
D1ERGO-Cam2	2.12	2.87	3.43	5.72	6.43	10.07	18.49	28.21	30.96	32.32
D1ERRG-Cam1	19.51	25.93	32.90	28.92	41.51	29.91	47.34	57.18	56.88	48.52
D1ERRG-Cam2	29.96	43.93	57.30	84.84	72.16	84.43	70.16	77.38	74.34	66.53
D1ERGR	4.90	6.99	7.53	12.17	16.09	27.44	40.84	55.94	56.59	41.48
D1ERTG	15.08	19.64	37.10	25.63	29.60	38.43	44.83	42.29	41.86	37.52
D1ERRC	2.47	3.57	5.06	8.33	14.23	23.26	47.45	81.02	92.87	92.57
D1ERCR	6.71	5.79	6.49	12.40	19.04	42.11	77.67	142.01	133.05	123.24
D1ERmR2G	9.84	13.56	17.92	22.72	25.01	50.53	82.87	107.33	115.81	107.91
D1ERGmR2	9.36	10.94	15.66	18.13	26.91	44.56	58.36	70.51	63.02	60.20
D1ERmR2C	1.99	2.26	2.74	3.68	5.21	11.83	18.93	30.48	31.09	40.35
D1ERCmR2	16.70	15.85	17.65	33.00	51.42	85.53	151.17	246.90	263.13	246.64

Figure 3. Column statistics of maximal ER Ca²⁺-release in HeLa cells transfected with individual GFP/RFP-based cameleons. Bars representing FRET change upon treatment with 100 µM histamine and 15 µM BHQ detected by D1ERRG-Cam1 (pink bar, n = 13), D1ERRG-Cam2 (purple bar, n = 10), D1ERGR (light grey bar, n = 4), D1ERTG (violet bar, n = 7), D1ERRC (light green bar, n = 10), D1ERCR (grey bar, n = 4), D1ERGmR2 (black bar, n = 11), D1ERmR2G (dark red bar, n = 6), D1ERmR2C (blue bar, n = 11) or D1ERCmR2 (red bar, n = 7). * P < 0.05 for D1ERCmR2 vs. all other indicators.

Figure 3. Column statistics of maximal ER Ca²⁺-release in HeLa cells transfected with individual GFP/RFP-based cameleons. Bars representing FRET change upon treatment with 100 µM histamine and 15 µM BHQ detected by D1ERRG-Cam1 (pink bar, n = 13), D1ERRG-Cam2 (purple bar, n = 10), D1ERGR (light grey bar, n = 4), D1ERTG (violet bar, n = 7), D1ERRC (light green bar, n = 10), D1ERCR (grey bar, n = 4), D1ERGmR2 (black bar, n = 11), D1ERmR2G (dark red bar, n = 6), D1ERmR2C (blue bar, n = 11) or D1ERCmR2 (red bar, n = 7). * P < 0.05 for D1ERCmR2 vs. all other indicators.

3.3. D1ERCmR2 Has a K_d of 200 µM in Situ and Respond to Ca²⁺ over a Broad Range

In order to verify whether or not GFP/RFP-based ER targeted cameleons have a K_d that is suitable for imaging ER Ca²⁺ dynamics we determined the K_d of D1ERCmR2 in situ. Therefore we used ionomycin and digitonin permebilized HeLa cells and titrated Ca²⁺ at various concentrations ranging from 1 µM to 10 mM. These experiments revealed a K_d of 215.9 (150.7–309.4) µM for D1ERCmR2 (Figure 4). Originally in vitro studies of D1ER revealed 2 K_ds, one at 0.8 µM and another at 60 µM [17,34]. However, in in situ experiments also using HeLa cells, the K_d of this probe was determined to be 220 µM [42] which is in agreement with the K_d that was found in this study for the red-shifted D1ERCmR2. Notably, the Hill slope of the Ca²⁺ concentration response curve of D1ERCmR2 was found to be 0.50 (0.39–0.60), which indicates that the red-shifted cameleon allows better detecting changes of ER Ca²⁺ over a broad range.

Figure 4. Determination of the K_d in permeabilized HeLa cells expressing D1ERCmR2. [Ca²⁺] was titrated to quantify the FRET ratio in percentage upon Ca²⁺-binding within D1ERCmR2 at 1 µM (n = 21), 3 µM (n = 21), 10 µM (n = 37), 30 µM (n = 37), 100 µM (n = 26), 300 µM (n = 43), 1 mM (n = 35), 2 mM (n = 17) and at 10 mM (n = 18).

Figure 4. Determination of the K_d in permeabilized HeLa cells expressing D1ERCmR2. [Ca²⁺] was titrated to quantify the FRET ratio in percentage upon Ca²⁺-binding within D1ERCmR2 at 1 µM (n = 21), 3 µM (n = 21), 10 µM (n = 37), 30 µM (n = 37), 100 µM (n = 26), 300 µM (n = 43), 1 mM (n = 35), 2 mM (n = 17) and at 10 mM (n = 18).

3.4. Dual Recordings of [Ca²⁺]_ER and [Ca²⁺]_Cyto in Single Fura-2 Loaded Cells Expressing D1ERCmR2

Because of the different spectral properties of the red-shifted ER targeted cameleon D1ERCmR2 compared with fura-2 (Fura2), the combination of both Ca²⁺ sensors allows simultaneous recordings of [Ca²⁺]_ER and [Ca²⁺]_Cyto in single individual cells. As expected, imaging of fura-2 loaded HeLa cells that express D1ERCmR2 revealed no fluorescence overlap of the respective fluorescence signals (Figure 5A). Cell treatment with a combination of the IP₃-generating agonist histamine and the SERCA inhibitor BHQ instantly reduced the FRET ratio signal of D1ERCmR2 and increased the fura-2 ratio (Figure 5B and Video S1), indicating that under these conditions ER Ca²⁺ is mobilized quickly, resulting in a fast elevation of the cytosolic Ca²⁺ concentration. As the cell was stimulated in the absence of extracellular Ca²⁺, the cytosolic Ca²⁺ elevation turned back to basal levels within one minute. The subsequent addition of Ca²⁺ upon the removal of histamine and BHQ resulted in a transient rise of [Ca²⁺]_Cyto due to SOCE. The simultaneous recording of ER Ca²⁺ revealed that during SOCE, the ER gets refilled with a delay of almost 10 s (Figure 5C). Moreover, this approach highlights that the cytosolic Ca²⁺ elevation stops when approximately 40% of the ER Ca²⁺ content is restored (Figure 5C,D). While these data are in line with the kinetics of STIM1 oligomerization and punctae formation [43] further experiments are necessary to investigate the temporal correlation of SOCE activation and termination with respective ER Ca²⁺ dynamics.

Figure 5. Dual visualization of fura-2 and D1ERCmR2 in single individual HeLa cells. (A) Confocal images of D1ERCmR2 expressing and fura-2 loaded HeLa cells. Subcellular structures of D1ERCmR2 were either illuminated at 473 or 515 nm and emissions were recorded at 525 (Clover, upper left panel) or 570 nm (mRuby2, upper right panel), respectively. Fura-2 was excited at 405 and emitted light was imaged at 460 nm (lower left panel). Images overlaid (lower right panel); (B) Representative curves of cytosolic and ER [Ca²⁺] in a single HeLa cell treated with 100 µM Histamine and 15 µM BHQ in a nominal Ca²⁺ free buffer and subsequent readdition of 2 mM Ca²⁺; (C) Zoom into event of SOCE reveals a delayed ER Ca²⁺ refilling; (D) Spatiotemporal correlation of [Ca²⁺]_Cyto and [Ca²⁺]_ER during SOCE in percentage of the maximum increase shown in panel C.

Figure 5. Dual visualization of fura-2 and D1ERCmR2 in single individual HeLa cells. (A) Confocal images of D1ERCmR2 expressing and fura-2 loaded HeLa cells. Subcellular structures of D1ERCmR2 were either illuminated at 473 or 515 nm and emissions were recorded at 525 (Clover, upper left panel) or 570 nm (mRuby2, upper right panel), respectively. Fura-2 was excited at 405 and emitted light was imaged at 460 nm (lower left panel). Images overlaid (lower right panel); (B) Representative curves of cytosolic and ER [Ca²⁺] in a single HeLa cell treated with 100 µM Histamine and 15 µM BHQ in a nominal Ca²⁺ free buffer and subsequent readdition of 2 mM Ca²⁺; (C) Zoom into event of SOCE reveals a delayed ER Ca²⁺ refilling; (D) Spatiotemporal correlation of [Ca²⁺]_Cyto and [Ca²⁺]_ER during SOCE in percentage of the maximum increase shown in panel C.

The approach of dual recordings of [Ca²⁺]_Cyto and [Ca²⁺]_ER was further tested using HEK-293 cells stably expressing the ryanodine receptor 2 (RYR2). RYR2 is referred to as cardiac Ca²⁺ release channel that makes HEK-293 cells (HEK E2) inducible to extracellular Ca²⁺ [35,44]. Ca²⁺ addition is known to activate store overload induced Ca²⁺ release (SOICR) in the HEK E2 clone. SOICR is frequently used to characterize the RYR2 channel activity using these heterologous expression system [35,44]. In simultaneous measurements using D1ERCmR2 as a sensor for [Ca²⁺]_ER and fura-2 as an indicator for [Ca²⁺]_Cyto, we observed cytosolic and ER Ca²⁺ oscillations in the HEK E2 clone (Figure 6A–C). This data show that the on and off kinetics of D1ERCmR2 are as fast to follow such SOICR-dependent oscillations and that this approach can be used to investigate the Ca²⁺-dependent activation of the RYR2 in more detail. Particularly, the simultaneous recording of [Ca²⁺]_Cyto and [Ca²⁺]_ER in this cell model should allow determining the ratio of [Ca²⁺]_Cyto/[Ca²⁺]_ER upon clusters of RyR2 get activated. However, fura-2 as a Ca²⁺ buffer might affect the frequency of SOICR under these conditions.

Figure 6. Time course of [Ca²⁺]_Cyto and [Ca²⁺]_ER in a single HEK E2 cell visualizing oscillatory traces in 1 mM Ca²⁺. (A) Single emission curves of fura-2 either excited at 340 nm (black curve) or 380 nm (blue curve); (B) Donor and FRET acceptor fluorescences of D1ERCmR2 at 560 nm (red curve) or 510 nm (green curve), respectively; (C) Overlay of ratio curves from fura-2 (blue curve) or D1ERCmR2 (red curve).

Figure 6. Time course of [Ca²⁺]_Cyto and [Ca²⁺]_ER in a single HEK E2 cell visualizing oscillatory traces in 1 mM Ca²⁺. (A) Single emission curves of fura-2 either excited at 340 nm (black curve) or 380 nm (blue curve); (B) Donor and FRET acceptor fluorescences of D1ERCmR2 at 560 nm (red curve) or 510 nm (green curve), respectively; (C) Overlay of ratio curves from fura-2 (blue curve) or D1ERCmR2 (red curve).

4. Conclusions

In this study, we show that GFP/OFP-based cameleons are impractical to monitor ER Ca²⁺ signals, while red-shifted ER targeted cameleons that consist of a GFP/RFP FRET pair are suitable to image ER Ca²⁺ dynamics in single cells. A cameleon, referred to as D1ERCmR2, which is based on the Clover/mRuby FRET pair, was shown to have a good signal to noise ratio, optimal K_d and Hill slope to record dynamic changes of ER Ca²⁺ in mammalian cells. Moreover, this red-shifted probe can be used simultaneously with fura-2 in order to correlate [Ca²⁺]_ER and [Ca²⁺]_Cyto in single individual cells. With this work, we expanded the palette of red-shifted organelle targeted Ca²⁺ probes that show less phototoxicity and should be compatible with most of the existing confocal and wide-field fluorescence imaging systems.