A green lifetime biosensor for calcium that remains bright over its full dynamic range

  1. Franka H van der Linden
  2. Stephen C Thornquist
  3. Rick M ter Beek
  4. Jelle Y Huijts
  5. Mark A Hink
  6. Theodorus W J Gadella
  7. Gaby Maimon
  8. Joachim Goedhart  Is a corresponding author
  1. Swammerdam Institute for Life Sciences, Section of Molecular Cytology, van Leeuwenhoek Centre for Advanced Microscopy, University of Amsterdam, Netherlands
  2. Laboratory of Integrative Brain Function and Howard Hughes Medical Institute, The Rockefeller University, United States
14 figures, 5 tables and 3 additional files

Figures

Figure 1 with 4 supplements
The excitation and emission peaks of G-Ca-FLITS are red-shifted relative to Tq-Ca-FLITS.

(A) The excitation and emission spectra show in lighter lines the calcium-free state (10 mM EGTA) and the darker lines the calcium-bound state (39 μM free Ca2+). The Tq-Ca-FLITS is shown as a reference and was reported before (van der Linden et al., 2021). Spectra are normalized to the maximum of the calcium-free state for G-Ca-FLITS. (B) The photo shows the fluorescence from tubes of purified biosensors when excited with UV (312 nm).

Figure 1—figure supplement 1
Sequence alignment of Tq-Ca-FLITS and G-Ca-FLITS.

Residues with yellow background highlight mutations in G-Ca-FLITS relative to the parental Tq-Ca-FLITS. The RS20 peptide is in blue, the FP in black, and calmodulin in orange. Linker residues are underlined.

Figure 1—figure supplement 2
Excitation and emission spectra of Tq-Ca-FLITS_T203Y and mTq2_T203Y.

The gray line indicates the calcium-free state (10 mM EGTA) and the black line the calcium-bound state (39 μM free Ca2+). Spectra are normalized to the top of the calcium-free state.

Figure 1—figure supplement 3
Quantum yield (QY) determination.

For each protein, the absorbance and emission spectra were measured of four dilutions (indicated by points). This was done in both presence (39 μM free Ca2+) and absence of calcium (10 mM EGTA) for the sensor variants. For rhodamine, we used eight dilutions on 2 days. The measured absorbance and the integrated emission are plotted here. Linear models (lines) are fitted through the measurements and forced through the origin. Gray bands indicate the 95% confidence interval of the models.

Figure 1—figure supplement 4
Determination of the extinction coefficient by unfolding of the proteins.

Absorbance was measured before (solid line) and after (dotted line) unfolding by 0.5 M NaOH and 2 M urea of the proteins G-Ca-FLITS, Tq-Ca-FLTIS_T203Y, and mTq2_T203Y. For both sensors, the determination was done in the presence (black, 39 μM free Ca2+) and absence (gray, 10 mM EGTA) of calcium. Spectra are corrected for the difference in dilution between the folded and unfolded state and normalized to the maximum of the unfolded spectrum.

Figure 2 with 1 supplement
Influence of pH on phase lifetime of the proteins.

Fluorescence lifetime of proteins diluted in pH buffer was measured (n = 3). In the case of G-Ca-FLITS, this was done in the presence (gray, 0.1 mM CaCl2) or absence of calcium (black, 5 mM EGTA). A smooth curve is fitted through the data using the loess method, using α = 0.4. The gray band indicates the 95% confidence interval of the smooth fit.

Figure 2—figure supplement 1
Influence of pH on modulation lifetime of the proteins.

Fluorescence lifetime of proteins diluted in pH buffer was measured (n = 3). In the case of G-Ca-FLITS, this was done in the presence (gray, 0.1 mM CaCl2) or absence of calcium (black, 5 mM EGTA). A smooth curve is fitted through the data using the loess method, using α = 0.4. The gray band indicated the 95% confidence interval of the smooth fit.

Figure 3 with 1 supplement
Calcium calibration of G-Ca-FLITS in vitro at 37°C.

(A) The measured fluorescence lifetime of protein isolate of G-Ca-FLITS in a range of calcium concentrations is plotted in a polar space (n = 3), with the color indicating the concentration. The measurements fall on a straight line (in gray) on the polar plot between the lowest and highest concentration (indicated by an X). (B) For each measurement, the fraction of sensors in the calcium-bound state (the low lifetime) is determined, taking the intensity contribution of the two extreme states into account. The fraction is plotted against the concentration of free calcium to obtain a calibration curve.

Figure 3—figure supplement 1
Calcium calibration of G-Ca-FLITS in vitro at 37°C and at RT.

(A) Measured fluorescent phase and (B) modulation lifetime at 37°C for a range of calcium concentrations. (C) The line fraction of sensors in the calcium-bound state (the low lifetime) is determined, without considering the intensity contributions of the states, KD = 234 nM and Hill coefficient = 1.5. (D) Measured fluorescent phase and (E) modulation lifetime at RT for a range of calcium concentrations. (F) The fluorescence lifetime at RT is plotted in a polar space, showing a straight line (in gray) between the average of the lowest and highest concentration (indicated by an X). (G) The line fraction is determined for each measurement at RT, with respect to the calcium-bound state (low lifetime state), KD = 337 nM and Hill coefficient = 1.7. (H) The true fraction of sensors in the calcium-bound state at RT is determined by considering the intensity contribution of both states. In each panel, the circles depict the individual measurements (n = 3), and the gray lines are the fitted calibration curves. In the polar plot, the concentration of free calcium is indicated by the color of the circles.

Brightness analysis of green fluorescent proteins and biosensors in HeLa cells using a co-expressed mScarlet-I.

The ratio GFP/RFP was determined 24 hr after transfection both in resting, unstimulated cells (pre) and cell stimulated with 5 μg/ml ionomycin and 5 mM calcium to reach saturation of the sensors (post). The ratios were normalized to cells expressing EGFP. Dots show individual cell data (n ≥ 99) and the large dot the average of each biological replicate (N = 2, except for EGFP; N = 3).

Figure 5 with 1 supplement
Comparison of the signal-to-noise ratio (SNR) of the phase lifetime of G-Ca-FLITS and Tq-Ca-FLITS in HeLa cells.

HeLa cells were measured in a resting state and after addition of 5 μg/ml ionomycin and 5 mM calcium. Measurements of individual cells are indicated by colored dots (n = 18 for G-Ca-FLITS, n = 9 for Tq-Ca-FLITS). Lines connect the measurements of the same cell. The average of all cells is shown in black. For each cell, 400 pixels were analyzed for the mean lifetime and intensity, and for the sd of the lifetime. SNR = mean lifetime/sd lifetime. Comparable phase lifetimes are found for all cells expressing a construct, but the SNR is lower for lower intensity cells and varies less between the two states for G-Ca-FLITS.

Figure 5—figure supplement 1
Comparison of the signal-to-noise ratio (SNR) of the modulation lifetime of G-Ca-FLITS and Tq-Ca-FLITS in HeLa cells.

HeLa cells were measured in a resting state and after addition of 5 μg/ml ionomycin and 5 mM calcium. Measurements of individual cells are indicated by colored dots (n = 18 for G-Ca-FLITS, n = 9 for Tq-Ca-FLITS), averages of all cells are indicated in black. For each cell, 400 pixels were analyzed for the mean lifetime and the sd of the lifetime. SNR = mean lifetime/sd lifetime. Comparable modulation lifetimes are found for all cells expressing a construct; however, the SNR is lower for lower intensity cells and varies less between the two states for G-Ca-FLITS.

Figure 6 with 2 supplements
Measuring calcium concentrations in HeLa cells and Blood Outgrowth Endothelial Cells (BOECs) using G-Ca-FLITS targeted to various organelles.

(A) Localization of G-Ca-FLITS in HeLa cells and (B) in BOECs. Images are taken with a ×63 magnification. (C) Measured calcium concentration in various organelles in HeLa cells and (D) BOECs. Measurements of single cells are indicated by circles, gray for unstimulated cells and black for cells after addition of 5 μg/ml ionomycin and 5 mM calcium. The mean of all cells is indicated by a black line. (E) Changing free calcium concentration after stimulation with 1 μM histamine or addition of 5 μg/ml ionomycin and 5 mM calcium in the cytosol and (F) in the mitochondria. Each line represents a single cell. Arrows indicate the moments of addition.

Figure 6—figure supplement 1
Confocal lifetime imaging of G-Ca-FLITS expressed in mitochondria of HeLa cells.

(A) Intensity and false color images of the median arrival time. The peripheral regions with a differential lifetime are indicated in blue, the cell bodies are indicated in orange. Images were taken with a PicoQuant TCSPC setup with ×60 magnification with a total integration time of ~5 min. (B) Two lifetime components (τ1 and τ2) determined from decay curves of the two types of areas. Each circle indicates either a single ‘peripheral area’ (blue) or a single ‘cell body’ (orange). The mean is indicated in black, including the standard deviation.

Figure 6—figure supplement 2
Aggregated FLIM data of measurements in organelles under resting conditions and after incubation with ionomycin.
Figure 7 with 1 supplement
Quantitative imaging of spontaneous calcium dynamics in mitochondria.

(A, B) FLIM images were acquired with a Leica Stellaris 8 every 60 s and at the end of the sequence 10 µM digitonin was added to obtain a maximal response. The lifetime images were converted to calcium concentration using the KD determined in vitro and the extremes from the lifetime image. False colors reflect the calcium concentration, according to the scale bar on the left. The size of the images is 132 µm x 132 µm. Calcium image at t = 1 (A) and t = 11 min (B), showing the three regions where calcium concentrations were quantified and displayed over time (C).

Figure 7—figure supplement 1
In vitro calibration of G-Ca-FLITS on the Leica Stellaris 8 at room temperature.

(A) Purified protein was added to calcium buffers and lifetimes images were acquired at room temperature. The fluorescence lifetime in a range of calcium concentrations is plotted in a polar space (n = 3), with the color indicating the concentration. The measurements fall on a straight line on the polar plot between the lowest and highest concentration. (B) For each measurement, the fraction of sensors in the calcium-bound state (the low lifetime) is determined, taking the intensity contribution of the two extreme states into account. The fraction is plotted against the concentration of free calcium to obtain a calibration curve. The solid line is a fit with KD = 0.483 µM and Hill coefficient 1.7.

Multiplex imaging of Tq-Ca-FLITS and G-Ca-FLITS.

FLIM images of HeLa cells co-transfected with 3xnls-Tq-Ca-FLITS and G-Ca-FLITS-3xnes were acquired with a Leica Stellaris 8 before (A) and after addition of 100 µM histamine (B). False colors reflect the lifetime (average arrival time), according to the scale bar on the right. For the cyan channel, the excitation was at 440 nm and detection at 450–490nm and for the green channel, the excitation was at 490 nm and detection at 500–540nm.

Figure 9 with 1 supplement
Two-photon time-correlated single photon counting (TCSPC) signals from G-Ca-FLITS and Tq-Ca-FLITS, and comparison to jGCAMP7f, in an intact Drosophila brain.

(A) Schematic of a head-fixed live Drosophila melanogaster female, imaged with standard saline for 2 min before perfusing a high [K+] saline variant over the brain. (B) Single example traces of flies expressing jGCaMP7f (pink, left), Tq-Ca-FLITS (turquoise, center), or G-Ca-FLITS (green, right) in the EPG neurons under the control of R60D05-Gal4. Top row: Tq-Ca-FLITS and G-Ca-FLITS show strong FLIM changes in response to elevated [K+] while jGCaMP7f shows very little. Bottom row: jGCaMP7f and Tq-Ca-FLITS exhibit large changes in fluorescence while G-Ca-FLITS becomes moderately dimmer. Gray region indicates time when high K+ saline was perfused. (C) Intensity traces from all flies. Vertical scalebars: 0.200 ns for G-Ca-FLITS and Tq-Ca-FLITS, 5 ΔF/F for jGCaMP7f. Horizontal scalebars: 1 min. Gray region indicates time when high K+ saline was perfused. (D) Aggregated responses, the thick lines represent the mean response of all measurements. (E) Fluorescence (left) and FLIM (right) images (72 x 72 µm) of example flies from panel B before [K+] elevation (baseline, 2 min average, top row) and for the 30 s surrounding the peak change in calcium (second row). Lifetime images are masked to show only automatically determined ‘foreground’ pixels. Third row: photon arrival time histograms for the data of the above rows. (F–H) Summary statistics for all flies (from panel C). (F) shows changes in fluorescence intensity, (G) shows absolute lifetime measurements, and (H) reflects the change in lifetime from baseline to plateau for each of the three indicators.

Figure 9—figure supplement 1
Similar brightness for G-Ca-FLITS and Tq-Ca-FLITS between 1-Photon Excitation (1PE) and 2-Photon Excitation (2PE).

Ni-NTA beads were loaded with purified Tq-Ca-FLITS or G-Ca-FLITS that are both equipped with a 6xHis tag. Images were acquired with either 1PE (460 nm) or 2PE (920 nm), and the intensities of the beads were quantified. The smaller dots are average intensity of individual beads, and the larger dot is the average. The overall higher intensity of Tq-Ca-FLITS is due to a higher concentration of protein loaded on the beads.

Appendix 1—figure 1
Influence of different components of a bacterial lysis protocol.

Emission spectra of bacterial lysates expressing Tq-Ca-FLITS are shown. Various concentrations of urea, DOC, and lysozyme were used in the lysis buffer. Each buffer also contained 50 mM Tris-HCl (pH 8.0). The influence of a freeze/thawing step is also tested, indicated by ‘frozen’. Each condition was measured once. The condition that was selected for the screening is indicated with a green box.

Appendix 1—figure 2
Influence of different detergents on a bacterial lysis protocol.

Emission spectra of bacterial lysates expressing Tq-Ca-FLITS are shown. Various concentrations of SDS and Triton X-100 were tested and compared to a buffer with 2% DOC and a buffer with 1% DOC and 1 M urea, the original bacterial lysis buffer (Danilevich et al., 2008). Each buffer also contained 50 mM Tris-HCl (pH 8.0). Each condition was measured three times.

Appendix 1—figure 3
Excitation (A) and emission spectra (B) of screened variants in the search for a red-shifted calcium sensor based on Tq-Ca-FLITS.

The spectra were measured in bacterial lysates, with addition of 0.1 mM CaCl2 (black line) or 9.5 mM EDTA (gray line). Sensor variants are indicated by their circular permutation and mutations in the FP in the sensor. For example, ‘cp146’ has from N- to C-terminus the following domains: calmodulin binding peptide M13, amino acids 146–283 of mTq2, a flexible GGSGG linker, amino acids 1–145 of mTq2, CaM (calmodulin). If no circular permutation position is indicated, the mutation displayed is done on Tq-Ca-FLITS. ‘T203Y’ indicates a T203Y mutation on the mTq2 in the sensor. The light blue shading is used as a reference for excitation and emission peaks of Turquoise.

Appendix 2—figure 1
Effective excitation and detected emission of G-Ca-FLITS for the indicated filters.

Excitation and emission spectra were recorded using isolated protein. Filter profiles and the detector sensitivity are provided on the manufacturers’ respective websites. The LED brightness was measured. (A) All normalized spectra related to excitation of G-Ca-FLITS are shown. The ‘effective excitation pulse’ (‘LED’ × ‘excitation filter’ × ‘reflection dichroic mirror’) is shown in green. The ‘effective excitation pulse’ is multiplied with the ‘excitation spectra’ to yield the ‘effective excitation’ of G-Ca-FLITS for each state (cyan and orange). The erel discussed in Appendix 2 is the integral of ‘effective excitation’ divided over the integral of ‘excitation pulse’. (B) All normalized spectra related to emission and detection of G-Ca-FLITS are shown. The ‘detected emission’ of G-Ca-FLITS was calculated for each state by multiplication of the following spectra: ‘transmission dichroic mirror’ × ‘emission filter’ × ‘detector sensitivity’ × ‘emission spectrum’. The Srel discussed in Appendix 2 is the integral of ‘detected emission’ divided over the integral of ‘emission spectra’ for each calcium state.

Author response image 1

Tables

Table 1
Lifetime and spectral screen of candidate sensors to create G-Ca-FLITS.

Sensor variants are indicated by their mutations with respect to Tq-Ca-FLITS. The calcium-bound state (sat) is measured in presence of 0.1 mM CaCl2 and the calcium-free state (apo) after addition of 9.5 mM EDTA.

Sensor variantExcitation max. (nm)*Emission max. (nm)*Phase lifetime (ns)
aposataposataposatChange
Tq-Ca-FLITS437, 455437, 456492, 502480, 5071.722.861.14
T203F4594585065071.712.190.48
T203HNo red shiftNo red shiftNo red shiftNo red shift2.822.900.08
T203Y461, 488459, 4785165143.082.20–0.88
I167F, T203Y460, shoulder 480460, 479512, shoulder 4905162.592.880.29
I167H, T203Y459, 478460, 4795145192.923.110.18
I167FNo red shiftNo red shiftNo red shiftNo red shift1.522.170.65
  1. *

    Emission and excitation maxima are only indicated if a red shift of the spectrum with respect to Tq-Ca-FLITS was observed.

  2. Phase lifetimes were measured at 37°C by FD-FLIM. The lifetime change is calculated as the lifetime in the calcium-bound state minus the calcium-free state.

Table 2
Lifetime contrast of improved green versions of Tq-Ca-FLITS.
Sensor variant*StateBacterial lysateResponse in HeLa cells
Phase lifetime (ns)Phase lifetime (ns)Modulation lifetime (ns)
MeanChangeMeansdChangeMeansdChangen
G-Ca-FLITS
(Tq-Ca-FLITS_T203Y-AD)
apo3.83–1.373.190.08–1.173.650.07–1.15314
sat2.452.020.062.500.08267
Tq-Ca-FLITS_T203Y-VSapo3.39–1.232.900.06–0.923.370.07–0.95150
sat2.151.980.072.420.08116
Tq-Ca-FLITS_T203Y-ASapo3.85–1.233.190.08–0.963.630.07–0.89306
sat2.622.230.062.740.06345
Tq-Ca-FLITS_T203Yapo3.34–1.132.950.08–0.823.420.08–0.85193
sat2.212.130.062.570.07201
  1. *

    ‘AD’ stands for the mutations V27A and N271D, ‘VS’ for V27 and N271S, and ‘AS’ for V27A and N271S.

  2. The fluorescence lifetime of bacterial lysates was measured at RT by FD-FLIM after addition of 0.1 mM CaCl2 and after addition of 9.5 mM EDTA. Mean phase lifetimes of the full field of view of the microscope are indicated.

  3. The fluorescence lifetime of HeLa cells expressing the different sensor variants was measured without stimulation and after addition of 5 μg/ml ionomycin and 5 mM calcium to the medium, measured at 37°C. The mean and standard deviation (sd) of the fluorescence lifetime of all cells (n) are indicated.

Table 3
Photophysical properties of G-Ca-FLITS and intermediate variant.

Apo – calcium-free state (10 mM EGTA), sat – calcium-bound state (39 μM free Ca2+), ε – extinction coefficient, QY – quantum yield with 95% confidence interval between curly brackets.

Stateλabs(nm)λem(nm)εmax(M–1 cm–1)QY {CI}(%)
G-Ca-FLITSapo47451529,30041.1 {40.5–41.7}
sat47651730,90025.9 {25.4–26.5}
Tq-Ca-FLITS_T203Yapo47051628,90043.8 {43.2–44.3}
sat47551830,90030.7 {30.4–31.1}
mTq2_T203Y47151427,40023.8 {23.5–24.2}
Table 4
Calcium calibration parameters of G-Ca-FLITS.

The calibration was performed with purified protein at two temperatures. The phase and modulation lifetime of the sensor are indicated for the two states of the sensor. apo – in calcium-free buffer (10 mM EGTA), sat – in buffer with 39 mM free CaCl2, τφ and τM – phase lifetime and modulation lifetime, KD – calcium affinity with 95% confidence interval indicated between curly brackets, n – Hill coefficient with 95% confidence interval indicated between curly brackets.

TemperatureStateτφ (ns)τM (ns)KD {CI} (nM)n {CI}
37°Capo2.983.50209
{206–211}
1.53
{1.51–1.56}
sat1.932.43
RTapo3.523.99339
{335–343}
1.67
{1.64–1.70}
sat2.032.57
Key resources table
Reagent type (species) or resourceDesignationSource or referenceIdentifiersAdditional information
Strain, strain background (Escherichia coli)E.cloni 10GLucigen60106-1Electrocompetent cells
Genetic reagent (D. melanogaster)UAS-Tq-Ca-FLITSThis papery[1]w[1118]; PBac{y[+mDint2]w[+mC]=20XUAS-IVS-Tq-Ca-FLITS}VK00027
Genetic reagent (D. melanogaster)UAS-G-Ca-FLITSThis papery[1]w[1118]; PBac{y[+mDint2]w[+mC]=20XUAS-IVS-G-Ca-FLITS}VK00027
Cell line (Homo sapiens)HeLaATCCCCL-2
Cell line (Homo sapiens)cbBOECJaap van Buul
Recombinant DNA reagentG-Ca-FLITS (plasmid)This paperhttps://www.addgene.org/191465/Cytoplasm targeted biosensor
Recombinant DNA reagent4xmts-G-Ca-FLITS (plasmid)This paperhttps://www.addgene.org/191462/Mitochondria targeted biosensor
Recombinant DNA reagentER-G-Ca-FLITS-KDEL (plasmid)This paperhttps://www.addgene.org/191464/ER targeted biosensor
Recombinant DNA reagent3xnls-G-Ca-FLITS (plasmid)This paperhttps://www.addgene.org/191463/Nucleus targeted biosensor
Chemical compound, drug Polyethylenimine, Linear, MW 25000, Transfection GradePolysciences23966
Chemical compound, drugIonomycinL C LaboratoriesI-68005 µg/ml
Commercial assay, kitCalcium Calibration Buffer Kit #1, zero and 10 mM CaEGTAThermo Fisher ScientificC3008MP
Software, algorithmRRStudio
Software, algorithmFIJIdoi:https://doi.org/10.1038/nmeth.2019https://github.com/fiji/fiji
Software, algorithmFIJIdoi:https://doi.org/10.1038/nmeth.2019https://github.com/fiji/fiji; Schindelin et al., 2025

Additional files

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Franka H van der Linden
  2. Stephen C Thornquist
  3. Rick M ter Beek
  4. Jelle Y Huijts
  5. Mark A Hink
  6. Theodorus W J Gadella
  7. Gaby Maimon
  8. Joachim Goedhart
(2025)
A green lifetime biosensor for calcium that remains bright over its full dynamic range
eLife 14:RP105086.
https://doi.org/10.7554/eLife.105086.3