Molecular biology

The plasmid vector pRSET-B (Invitrogen) was used for the library generation, bacterial screenings and protein production. pcDNA 3.1+ (Invitrogen) and pCAGIG (Addgene, USA) were the expression vectors used for the studies in mammalian cells. All constructs were cloned using the homology-based SLiCE methodology⁴¹. Briefly, fragments were amplified by PCR containing 20 nucleotides overhangs overlapping with the required upstream and downstream sequences. SLiCE extract was prepared as suggested by the authors and 10 µl reactions were used with a vector/insert ratio of 1:1. E. coli XL1 Blue cells (Invitrogen) were used for all cloning steps and library screenings. Error-prone PCR mutagenesis was performed using the GeneMorph II Random mutagenesis kit (Agilent, Germany). Standard PCR reactions for cloning or site-mutagenesis were done using Herculase II Fusion polymerase (Agilent, Germany). All primers were ordered from Eurofins Genomics, Germany.

Directed evolution

The DNA libraries were incorporated into E. coli XL1 Blue cells and transferred onto LB-Agar plates supplemented with ampicillin (Roth), obtaining 500–800 colonies per plate. Depending on the size of the library, 20–50 plates were screened in each round. After an overnight incubation at 37 °C, LB-agar plates containing the libraries were further incubated at 4 °C during 24 h prior imaging to favor protein folding. Using a home-built screening platform⁴², the plates were imaged before and after the addition of 10 ml of screening buffer (30 mM MOPS, 100 mM KCl, 50 mM CaCl₂, 0.5%, agar pH 7.2) maintained at 42 °C during the screening. Images were acquired and analyzed using a customized python routine⁴². In general, after each round of screening, the selected indicators were expressed and purified for further characterizations. The clones were filtered first according to its response to calcium, then its extinction coefficient, and finally if a suitable candidate was identified, its response was evaluated in mammalian cells. Calibrations and comparisons were made using GCaMP6s as a control indicator. Once an acceptable response to calcium was obtained ( > 3000 %), new rounds of screenings were done focusing on improving protein expression, adjusting the kinetic performance and the calcium-binding affinity. Further details on the strategy for directed evolution of GreenT-ECs are given in Supplementary Note.

Protein purification

Proteins with histidine-tags were expressed in E. coli BL21 (Invitrogen) overnight at 37 °C in 50 ml auto-inductive LB (LB supplemented with 0.05% D-(+)- glucose (w/v), 0.2% lactose (w/v), 0.6% glycerol (v/v)). Bacteria were harvested by centrifugation (4 °C, 10 min, 6000 x g) and re-suspended in 10 ml Resuspension buffer (20 mM Na₂PO₄, 300 mM NaCl, 20 mM imidazole) (Sigma Aldrich) supplemented with protease inhibitors (4 μM PMSF, 20 μg/ml Pepstatin A, 4 μg/ml Leupeptin) (Sigma Aldrich), 5 μg/mL DNase and 10 μg/ml RNAse (Sigma Aldrich). Bacteria were first lysed physically through sonication on ice for 7 min 0.8 cycle, 0.8 power output (Bandelin Sonoplus). Insoluble components were pelleted through centrifugation (4 °C, 30 min at 20,000x g). For purification, the supernatant was incubated with 150 μl 6% (v/v) Nickel-IDA agarose bead suspension (Jena Bioscience) for 2 h at 4 °C under mild agitation. Agarose beads were collected in 1 ml propylene gravity flow columns (Qiagen) and washed with 10 ml Resuspension buffer. The proteins were collected using 700 μl elution buffer (20 mM Na₂PO₄, 300 mM NaCl, 300 mM imidazole) (Sigma Aldrich) and dialyzed against MOPS buffer (30 mM MOPS, 100 mM KCl, pH 7.2).

In vitro spectroscopy

The fluorescence change (ΔF/F₀) of the indicators was determined in 96-well plates using a fluorescence plate reader (Tecan). The Ca²⁺ free fluorescence (F₀) was measured in MOPS buffer supplemented with EGTA 0.4 mM. The fluorescence of the indicator corresponding to the Ca²⁺ bound state was measured in MOPS buffer supplemented with 100 mM Ca²⁺ and 1 mM Mg²⁺.

The molar extinction coefficient (EC) was determined through the absorption of the denatured chromophore at 452 nm (extinction coefficient 44 mM⁻¹cm⁻¹). Proteins were prepared in MOPS buffer supplemented with 60 mM CaCl₂ and the absorbance spectrum was acquired before and after the addition of NaOH to a final concentration of 0.025 M.

The quantum yield of new variants was determined relative to mNeonGreen by using the slope method. First, the absorption and emission spectra of three serial 1:2 dilutions were acquired in the same cuvette. Then, the integrated emission spectrum was plotted against the maximum absorption and the slope was determined. For mNeonGreen a quantum yield of 0.8 was adopted²⁰. The Ca²⁺ affinity of the indicators was determined using MOPS buffer supplemented with 10 mM EGTA, 1 mM Mg²⁺ and increasing concentrations of Ca²⁺ as previously described⁴³. Dissociation constant (Kd) values were determined by plotting the log10 values of the [Ca²⁺] free concentrations in mol/l against the corresponding ΔF/F₀ values (normalized to ΔF/F₀ at 39.8 μM Ca²⁺), and fitting a sigmoidal curve to the plot. Prism was used for data analysis.

To determine the pKa of a fluorescent protein (indicating its pH stability), a series of MOPS/MES buffered solutions supplemented with 100 mM Ca²⁺ were prepared, with pH values adjusted in 0.5 pH steps from pH 5.5 to pH 8.5 using NaOH and HCl. In a bottom 96 well plate, triplicates of 200 μl of buffer containing 0.5–1 μM of protein were prepared for each pH value and incubated for 10 min at room temperature. Subsequently, all emission spectra were recorded. To determine the pKa value, the relative fluorescence values at the protein’s emission maximum were plotted against the pH values and a sigmoidal fit was applied.

To determine the kinetic rates of the calcium indicators, a Cary Eclipse fluorescence spectrophotometer (Varian) fitted with an RX pneumatic drive unit (Applied Photophysics) was used. For obtaining the macroscopic off-rate constant (Koff), two stock solutions were prepared as follows: a calcium-saturated indicator solution (30 mM MOPS, 50 mM CaCl₂, 2 mM MgCl₂, 100 mM KCl, ∼ 0.2–1 μM indicator, pH 7.2) and a BAPTA solution (30 mM MOPS, 100 mM KCl, 100 mM BAPTA, pH 7.2). The stopped-flow experiment was carried out at room temperature (∼23 °C) and the two solutions were mixed with an injection pressure of 3.5 bar. Excitation was set to 480 nm and emission was detected at 520 nm. The acquisition time was set to 12.5 ms, duration to >10 s and mixing volume to 400 μl with a mixing dead time of the instrument of 8 ms. The decay time (τ, s) was determined by fitting with a double-exponential curve to the fluorescence response using Prism. Macroscopic on-rate kinetics (K_obs) were obtained by mixing the calcium-free buffer containing the protein (30 mM MOPS, 100 mM KCl, 1 mM MgCl₂, ∼ 0.2–1 μM indicator, pH 7.2) and solutions containing increasing concentrations of CaCl₂ (30 mM MOPS, 100 mM KCl, 0.1–100 mM CaCl₂, 1 mM MgCl₂, ∼ 0.2–1 μM indicator, pH 7.2). Concentrations of free calcium were calculated using WEBMAXC STANDARD server.

Crystallization and structure determination

Crystals of the GreenT-EC intermediate variant named NRS 1.2 were formed using the sitting drop method. The precipitant solution was 0.2 M ammonium sulfate with 30% w/v PEG 400. Droplet conditions were: 0.4 µL total volume, 200 nL protein + 200 nL precipitant. The crystals were cryoprotected with addition of 30% of ethylene glycol and flash-frozen in liquid nitrogen. X-ray data sets were recorded on the 10SA (PX II) beamline at the Paul Scherrer Institute (Villigen, Switzerland) at wavelength of 1.0 Å using a Dectris Eiger 16 M detector with the crystals maintained at 100 K by a cryocooler. Diffraction data were integrated using XDS (BUILT = 20220220) and scaled and merged using AIMLESS (0.7.7); data collection statistics are summarized in Supplementary Table 1. Initially the NRS 1.2 data set was automatically processed at the beamline to 1.3 Å resolution and a structure solution was automatically obtained by molecular replacement using pdb 5MWC as template. The map was of sufficient quality to enable 90 % of the residues expected for NRS 1.2 to be automatically fitted using Phenix autobuild. The model was finalized by manual rebuilding in COOT2 (0.9.6) and refined using in Phenix refine (1.19.2).

Cell lines and tissue culture

HeLa and HEK 293 T (Invitrogen) cells were grown in high glucose Dulbecco’s Modified Eagle Medium with high glucose, pyruvate (Gibco) supplemented with 10% fetal bovine serum (FBS, Gibco), 29.2 mg/ml of L-glutamine, 10,000 units of penicillin and 10,000 µg streptomycin at 37 °C with 5% CO₂.

Evaluation of indicator surface display

Different combinations of signal peptides and membrane domains were evaluated in HeLa cells. The cells were seeded in 35 mm glass-bottom dishes (Mat-Tek) pre-coated with Poly-L-lysine (Sigma) and transfected with the different constructs one day before imaging, according to the manufacturer instructions (Lipofectamine 3000). Hoechst staining was performed immediately before imaging using a final concentration of 2 µg/mL for 20 min. The imaging buffer used for these experiments was Hank Balanced Salt Solution (HBSS) supplemented with 1 mM MgCl₂ and 3 mM CaCl₂. Images were acquired in a Leica Stellaris5 microscope with a 60x oil-immersed objective. Different export signal peptides and transmembrane domain were evaluated: IβCE (from Integrin-β of C. elegans)^44,45, Igk (from human CH29 light chain)^44,46, Nrxn1 (mouse Neurexin I first 63 aa)^44,46, PPA (N-terminal 24 aa of mouse preproacrosin signal peptide)²⁵, PDGFRB (Homo sapiens platelet derived growth factor receptor beta)⁴⁷, GPI (mouse Thy-1 glycosylphosphatidylinositol) anchoring domain²⁵, NXN (rat neurexin-1β) and NLG (rat neuroligin-1)^44,46.

Evaluation of reference proteins using Igk-PDGFR surface localized GreenT-EC

HeLa cells were seeded in 35 mm glass-bottom dishes (Mat-Tek) and transfected with the different constructs according to the manufacturer instructions (Lipofectamine 3000). The imaging buffer used for these experiments was Hank Balanced Salt Solution (HBSS) supplemented with 1 mM MgCl₂ and 3 mM CaCl₂. Images were acquired in a Nikon Spinning Disk confocal microscope with a 20x oil-immersed objective. In each case, the reference proteins mCarmine⁴², mScarlet⁴⁸, mCyRFP1²⁴, mCerulean3⁴⁹ and mTurquoise2⁵⁰ were inserted in the C-terminal extreme of the PDGFR domain, separated by a short linker. The mCyRFP1 referenced construct was compared to a cytosolic version (not bearing any export signal or transmembrane domain). In this case, mCyRFP1 was cloned at the N-terminus of GreenT-EC or GCaMP6f using a 20 aa hydrophobic flexible linker as previously described⁵¹. Cells were permeabilized using ionomycin (2.5 µM), which allowed the entry of extracellular calcium (1.5 mM), or stimulated with Histamine (200 µM) to induce a physiological increase in cytosolic calcium.

Affinity titrations on the surface of HEK 293 T cells

HEK293T cells were seeded on day one into 35 mm glass-bottom dishes (Mat-Tek) pre-coated with Poly-L-lysine (Sigma). On day three, the cells were transfected using 1 µg of the plasmids coding for the different variants according to the manufacturer instructions (Lipofectamine 3000). For each variant, three dishes were prepared. On day four, cells were washed with buffer MOPS 30 mM KCl 100 mM, EGTA 0.4 mM, pH 7.2 during 2 min. Then 3 ml of buffer MOPS 30 mM KCl 100 mM, pH 7.2 was added and cells were imaged in a Leica SP8 confocal microscope using a 63x water-immersed objective. For the titration, 80–100 µl of CaCl₂ stock solutions were added drop-wise to the dishes during time-lapsed imaging. Normally, images were acquired every 1 min.

Flow cytometry

HeLa cells were seeded in 35 mm dishes (Corning) and transfected with the different constructs according to the manufacturer instructions (Lipofectamine 3000). Twenty-four hours after transfection, cells were detached with Versene solution (Gibco, ThermoFisher) and collected in HBSS supplemented with CaCl₂ 3 mM to inactivate the detachment agent. After centrifugation at 1000 g during 5 min, cells were resuspended in 1.5 ml MOPS buffer supplemented with 3 mM CaCl₂ / 1 mM MgCl₂, and transferred to 5 ml round bottom tubes with cell strainer snap cap (Corning). All samples were evaluated in an Attune NxT Analyzer cytometer (ThermoFisher Scientific) before and after the addition of EGTA 5 mM. At least 100.000 events were recorded for each condition and further analyzed using FlowJo v10.8.1 (FlowJo, LLC). Events were first gated by forward and side scattering for selecting cells and excluding doublets as shown in the figures. Attune BL1-A channel (Excitation 488 nm – Emission 530/30 nm) was used for monitoring GreenT-EC, VL1-A (Excitation 405 nm – Emission 440/50 nm) for mTurquoise2 and mCerulean3 constructs, YL2-A (Excitation 561 nm – Emission 620/15 nm) for mScarlet, and YL3-A (Excitation 561 nm – Emission 695/40 nm) for mCarmine. Mean fluorescence values were used to calculate the response as: (F_Ca2+-F_EGTA)/F_EGTA. Data was plotted using GraphPad Prism and both ungated and gated full populations are submitted separately in an excel file.

Image processing and ratioing

The general processing procedure for mammalian cell titrations and zebrafish experiments consisted in: i) Noise reduction (typically using a median or Gaussian filter of 2 pixels), ii) Thresholding saturated pixels and noise/cytoplasm signals. During this process, the original image was divided by the Threshold mask to assign N/A values to the pixels that were thresholded out. This avoids problems arising from averaging cero-value pixels during the ROI analysis. iii) Finally, the processed GreenT-EC and mCyRFP1 channels were used to obtain a ratiometric image GreenT-EC/mCyRFP1. This method allowed us to monitor the ratio signals in the membrane with minor interference from background or low-intensity cytosolic signals. ImageJ macro routines were generated to perform the image processing, ROI detection/drawing, measurement of individual channels and ratio, and exporting of the results. The data analysis was done in GraphPad. Images were displayed with white background in cases where high contrast was required to clearly visualize the different structures⁵². The ImageJ scripts used can be found in the Source Data file.

For brain slices and dissociated neuronal culture experiments, all images were processed and analyzed using ImageJ software (NIH). For time-lapse 2 P acquisitions, bleach correction based on an exponential fit provided by ImageJ was performed. Since a reference fluorescent protein was not included in these experiments, no rationing was performed.

Primary neuronal cell culture

All experiments were performed in accordance with the European directive on the protection of animals used for scientific purposes (2010/63/EU). Banker culture of hippocampal neurons were prepared from 18 day pregnant embryonic Sprague-Dawley rats as described previously⁵³. Briefly, hippocampi were dissected in HBSS containing Penicillin-Streptomycin (PS) and HEPES and dissociated with Trypsin-EDTA/PS/HEPES. Neurons were plated in minimum essential medium supplemented with 10% horse serum on coverslips coated with 1 mg/ml poly-L-lysine in 60 mm petri dishes at a density of 250000 cells per dish. Following neuronal attachment, the coverslips were flipped onto 60-mm dishes containing a glial cell layer in Neurobasal-Plus medium supplemented with GlutaMAX (GIBCO, #35050-038) and B27-Plus Neuronal Supplement (GIBCO, A3653401). Cells were maintained at 37 °C with 5% CO₂ for 13–15 days. Neurons were transfected at DIV 7–9 using the calcium-phosphate co-precipitation method. Per dish, precipitates containing 10 µg plasmid DNA of GreenT-EC were prepared using the following solutions: TE (1 mM Tris–HCl pH 7.3, 250 mM EDTA pH 8), CaCl₂ (2.5 mM CaCl₂ in 10 mM HEPES, pH 7.2), 2 x HEPES-buffered saline (HEBS; 11 mM D-Glucose, 42 mM HEPES, 10 mM KCl, 270 mM NaCl and 1.4 mM Na₂HPO₄.2H₂O, pH 7.2). Coverslips containing neurons were moved to 12 well multi-well plates containing 450 µl/well of conditioned culture medium. The 50 µl precipitate solution was added to each well, in the presence of 2 mM kynurenic acid (Sigma-Aldrich #K3375) and incubated for 1h-1h30 at 37 °C. Then, cells were washed with nonsupplemented Neurobasal medium containing 2 mM kynurenic acid for 20 min and moved back to their original culture dish. Cells were imaged at DIV 13–15.

Organotypic hippocampal slice culture

Mice were housed under a 12 h light/12 h dark cycle at 20–22 °C with ad libitum access to food and water in the animal facility of the Interdisciplinary Institute for Neuroscience (University of Bordeaux/CNRS), and monitored daily by trained staff. All animals used were free of any disease or infection at the time of experiments. Pregnant females and females with litters were kept in cages with one male. We did not distinguish between males and females among the perinatal pups used for organotypic cultures, as potential anatomical and/or physiological differences between the two sexes were considered irrelevant in the context of this study. C57Bl/6 J wild-type mice were used for all brain slice experiments in this study. Brain organotypic cultures were prepared according to the method described previously⁵⁴. Briefly, hippocampal slices were obtained from postnatal P5-7 old C57BI/6 J mouse pups. The animals were quickly decapitated and hippocampi were dissected out in cold GBSS (ThermoScientific) with 10 mM Glucose (VWR) and 2 mM kynurenic acid (Sigma-Aldrich). They were sliced on a McIlwain tissue chopper to generate coronal brain slices of 350 µm thickness. After 30 min of incubation at 4 °C, the slices were transferred on sterilized hydrophilic polytetrafluoroethylene (PTFE) membrane (FHLC04700; Sigma-Aldrich, 0.45 µm pore size) pieces, which were placed on top of cell culture inserts (PICMORG50; Sigma-Aldrich 0.4 µm pore size). The inserts were held in a 6-well plate filled with 1 ml of medium (50% Basal medium eagle (BME), 25% Hank’s Balanced Salt Solution (HBSS, pH 7.2), 25% Horse Serum, 11.2 mmol/L glucose and 20 mM glutamine; all from Thermofisher) and cultured up to 14 days at 35 °C/5% CO₂. Culture medium was replaced every two days.

AAV production and injections

In order to introduce the GreenT-EC plasmid to neurons in organotypic brain slices, the GreenT-EC coding sequence was first inserted into a specific viral plasmid (Addgene #50477) with a neuron-specific promotor CaMKII. After testing the plasmid for its correct size and integrity via restriction enzyme digestion, the construct was sent to in-house virus production facility, which provided us with a construct of GreenT-EC integrated into AAV9 particles. The viral particles were injected into the hippocampal slices via microinjections using a glass pipette connected to Picospritzer (Parker Hannifin). Briefly, the virus was injected via a pipette positioned into the CA1 area of the slice by brief pressure pulses (30 ms; 15 psi). The virus was injected at least 7 days prior to the experiments.

STED and confocal microscopy of hippocampal neurons

We used a custom-built STED/confocal setup⁵⁵ constructed around an inverted microscope body (DMI 6000 CS, Leica Microsystems) which was equipped with a TIRF oil objective (x100, 1.47 NA, HXC APO, Leica Microsystems) and a heating box (Cube and Box, Life Imaging Services) to maintain a stable temperature of 32 °C. A pulsed-laser (PDL 800-D, PicoQuant) was used to deliver excitation pulses at 488 nm and a de-excitation laser (Onefive Katana 06 HP, NKT Photonics) operating at 594 nm was used to generate the STED light pulses. The STED beam was profiled to a donut shape using a spatial light modulator (Easy3D Module, Abberior Instruments). Image acquisition was controlled by the Inspector software (Abberior Instruments). The spatial resolution of the microscope was 175 nm (x-y) and 450 nm (z) in confocal mode and 60 nm (x-y) and 160 nm (z) in STED mode.

Confocal/-STED imaging of hippocampal neurons in dissociated cultures and organotypic brain slices

For imaging, either slices or dissociated neurons were transferred on their glass coverslip to an imaging chamber and immersed in an imaging medium (in case of slices: artificial cerebrospinal fluid, ACSF; consisted of (in mM) 119 NaCl, 2.5 KCl, 1.3 MgSO₄, 1 NaH₂PO₄ x 2H2O, 1.5 CaCl₂* x 2H₂O, 20 D-Glucose x H₂O and 10 HEPES (all from Sigma Aldrich); 280 mOsm; pH 7.4; in case of dissociated neurons: Tyrode solution; consisted of (in mM) 10 D-Glucose x H₂O, 100 NaCl, 5 KCl, 2 MgCl₂ x 6H₂O, 25 HEPES, 2 CaCl₂* x 2H₂O). Confocal time-lapse images of dissociated neurons had a field of view: 50 µm x 50 µm. STED single-plane images of organotypic brain slices were either 100 µm x 100 µm or 50 µm x 50 µm with a pixel size of 48,6 nm, 0.3 ms dwell time. The excitation power was 0.5 µW (measured before the objective) and the STED power was 30 mW. *The concentration of calcium in the solutions varied depending on the experiment.

Two-photon microscopy

We used a commercial two-photon microscope (Prairie Technologies) and 40X water immersion objective (NA 1.0; Plan-Apochromat, Zeiss). For GreenT-EC excitation, a two-photon laser (Ti:sapphire, Chameleon Ultra II; Coherent) was tuned to 850 nm; with laser power ranging from 10–25 mW in the focal plane. The fluorescence signal was collected in a nondescanned manner by PMT detectors. The imaging parameters were adjusted using the commercial software provided by Prairie (Prairie View). For image acquisition, individual slices were transferred to a submerged recording chamber filled with HEPES-based ACSF solution (in mM): 119 NaCl, 2.5 KCl, 1.3 MgSO₄, 1 NaH₂PO₄ x 2H₂O, 20 D-Glucose x H₂O, 10 HEPES and varying CaCl₂ concentrations (all from Sigma Aldrich); 300 mOsm; pH 7.4. For most of the experiments, we acquired time-lapse images of single planes (126 µm x 126 µm) every 1 s for a total number of 40 repetitions. If the imaging parameters differed, it is mentioned in the respective figure legend.

Calcium puffing experiments

In order to locally apply calcium solutions, we inserted a glass micropipette into the slice expressing GreenT-EC sensor. By injecting brief high-pressure pulses (10–100 ms, 15 psi) via a Picospritzer (Parker Hannifin), ACSF solutions of varying calcium concentrations were locally delivered into the region of interest. To avoid pressure-induced z-drift, the injection parameters, pulse duration and pressure level, were optimized.

Electrophysiology

Schaffer collateral fibers in hippocampal slices were electrically stimulated and evoked field excitatory postsynaptic potentials (fEPSP) were recorded in the stratum radiatum of hippocampal CA1. Two glass micro-electrodes (tip resistance 5-6 MΩ) for stimulation and recording were filled with ACSF and carefully positioned in the slice and placed at depths where imaging was performed. The current pulses (5–15 pulses, 0.2-0.3 ms in duration) were delivered via the stimulating electrode from a stimulus isolator (AMPI; Science Products). The stimulus strength varied between 10–40 µA. Field potentials were recorded using a patch clamp amplifier (Multiclamp 700B; Molecular Devices).

To block calcium extrusion from cells, we applied two calcium-pump blockers: 5 µM sodium-orthovanadate (SOV, Sigma Aldrich) and 50 µM benzamil hydrochloride hydrate (BHH, Sigma Aldrich) dissolved in HEPES-based ACSF. To block voltage-gated sodium channels, we applied 1 µM tetrodotoxin (TTX, Hello Bio). Inhibition of voltage-gated calcium channels was done using 100 µM CdCl₂ (Sigma Aldrich). Finally, 20 µM cyanquixaline (CNQX, Sigma Aldrich) was used to block AMPA receptors.

Zebrafish strains and husbandry

Transgenic zebrafish derived from the outbred AB strain were used in all experiments. Zebrafish were raised under standard conditions at 28 °C. Animals were chosen at random for all experiments. Zebrafish husbandry and experiments with all transgenic lines were performed under standard conditions as per the Federation of European Laboratory Animal Science Associations (FELASA) guidelines⁵⁶, and in accordance with institutional (Université Libre de Bruxelles (ULB)) and national ethical and animal welfare guidelines and regulation, which were approved by the ethical committee for animal welfare (CEBEA) from the Université Libre de Bruxelles (protocols 578N-579N). To generate the actb2:GreenT-EC construct, the vector containing 9.8 kb of zebrafish b-actin2 (actb2) promoter²⁶ was digested by SpeI/NotI and GreenT-EC cloned using the previously mentioned SLiCE methodology⁴¹. Transgenics were generated using the I-SceI meganuclease system. Two founders were isolated and screened for the strength of green and red fluorescence in the body. The line with brighter expression was used to perform all experiments, and was designated Tg(actb2:GreenT-EC)^ulb18.

For experiments, transgenic zebrafish larvae, Tg(actb2:GreenT-EC), were obtained from outcross of the transgenic adult animal to AB wild-type animals. Larvae between 2 and 6 dpf (days post-fertilization) were used in all experiments. Zebrafish have indeterminate growth with the rate varying with parameters such as fish density, so larvae were kept at a density of 1 larvae per 1 ml of E2 medium (7.5 mM NaCl, 0.25 mM KCl, 0.5 mM MgSO₄, 75 mM KH₂PO₄, 25 mM Na₂HPO₄, 0.35 mM NaHCO₃) supplemented with either 0.3, 0.03, 2 or 10 mM CaCl₂, 0.5 mg/L methylene blue, pH 7.4 since birth unless specified otherwise. Anesthesia was administered in E2 medium using 0.02% pH 7.0 tricaine methanesulfonate (MS-222; E10521; Sigma-Aldrich, Darmstadt, Germany).

Zebrafish confocal image acquisition

Animals were anesthetized in 0.02 % tricaine methanesulfonate (MS-222; E10521; Sigma-Aldrich, Darmstadt, Germany) and mounted in 1% Low-Melt Agarose (50080; Lonza, Basel, Switzerland) and imaged on a glass-bottomed FluoroDish (FD3510–100; World Precision Instruments (WPI), Sarasota, Florida) using a LSM 780 confocal microscope (Zeiss). Finfold and muscle were imaged using a 40x/1.1 N.A. water correction lens. Imaging frame was set at 1024 × 1024, and the distance between confocal planes was set up at 3 µm for Z-stack cover, on average, a thickness of 60 µm. Samples were excited with 488 nm laser and fluorescence was collected in the two channels simultaneously using detector width of 493–530 nm for GreenT-EC and 550–740 nm for mCyRFP1.

Zebrafish pharmacological treatments

All compounds for treating embryos were dissolved in DMSO according to manufacturer, diluted in E2 embryo medium (0.3 mM or 0.03 mM Ca²⁺), changed daily and embryos treated by immersion. The compounds, and concentrations used, with catalogue numbers were: EGTA (Roth), 0.3 mM for 10 min; Calhex 231 hydrochloride (SML0668-Sigma), 5 or 10 µM for 48 h (from 2 to 4 dpf); Calcitriol (C3078-TCI) 2.5 µM for 24 h (3 to 4dpf). Dafadine A (HY-16670-MedChemExpress) was applied at 12.5 µM from 3 to 24 h postfertilization.

Time-lapse imaging of zebrafish

For EGTA time-lapse imaging, 4 dpf larvae were immersed in 5 mM EGTA solution, followed by embedding in 1% low-melt agarose. Subsequently, the solidified agarose was covered with 1 ml of 5 mM EGTA solution.

Statistics and reproducibility

Representative pictures were included only after systematic repetitions along different experimental replicates, and this can be evidenced from the control groups, acquired multiple times and for every drug treatment or condition, both for zebrafish and brain slices experiments. In the case of mammalian cell images, they were supported by biological replicates and additional evidences using flow cytometry. In the case of purified proteins, during the development of the project we observed that the measurements of the biophysical parameters were highly reproducible among different protein preparations. In addition, we always include controls with published indicators or fluorescent protein to confirm stable values over time, and to monitor the condition of solutions and equipments.

For experiments with neuronal cultures and brain slices, the size and type of individual samples “n” (number of dishes or slice preparations) for given experiments, is indicated and specified in the figure legends. Only for the experiment with CdCl₂ n refers to the total optical windows studied (two slices, on tree different positions for each one, were analyzed in this case). In the case of zebrafish experiments, for each animal all cells visible in the imaged plane (at least 4 cells/region of interest) were analyzed for the different tissues (notochord and ventral/dorsal muscle). The mean ratio GreenT-EC/mCyRFP1 per animal was then plotted. All experiments included between 4 and 6 animals per group and the precise number is indicated in the figure legends.

Statistical analyses were performed using Graphpad Prism software. Data was tested for normality using the Saphiro-Wilk test. Treated groups were compared with control animals using a two-tailed unpaired t-test to obtain the statistical significance. For multiple comparisons, One-way ANOVA was used followed by a Tukey’s test. Asterisks in figures indicate p values as follows: * p < 0.05, ** p < 0.01, *** p < 0.001 and **** p < 0.0001.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.