Application Note

Comparison of the Ca2+-Sensitive Dyes Fluo-3 and Fluo-4 Used with the FLIPR® Fluorometric Imaging Plate Reader System

Download PDF    


In the FLIPR® System, fluorescence-based measurement of intracellular calcium typically involves the use of the visible spectrum excitable dyes fluo-3, Calcium Green-1 or Oregon Green 488 BAPTA-1. These dyes offer advantages over UVexcitable dyes, including reduced interference from samples and plastic autofluorescence and less cellular photodamage.

In addition to those mentioned above, Molecular Probes (Eugene, Oregon) offers an array of visible light-excitable calcium indicators. They differ mainly in their dissociation constant for calcium and in their excitation and emission spectra. However, there is a continued search for dyes with higher sensitivity and lower toxicity. Molecular Probes recently reported the development of a fluo-3 analogfluo-4. The only structural difference between the two dyes are two chlorine atoms in fluo-3 that are replaced by two fluorine atoms in fluo-4. The described beneÞts of fluo-4 are greater fluorescence output, faster cell loading, and equivalent fluorescent signal at lower loading concentrations (BioProbes 28, May 1998).

This application note describes a direct comparison of results obtained using fluo- 3 and fluo-4, loaded into M1-transfected CHO-cells. The FLIPR System was used to measure the calcium response to carbachol activation.

Materials and Methods

Cells and Culture Conditions

CHO-M1 cells, stably transfected with the muscarinic acetylcholine-receptor (M1), were obtained from ATCC (M1 WT3 - CRL1984). The cells were cultured in Ham’s F-12 media supplemented with 10% FBS, 2 mM glutamine, Pen/Strep, and 50 µg/ ml Geneticin. Cells were plated one day before the assay on black 96-well microplates with clear bottoms (Costar) at a concentration of 50,000 cells/well (100 µl/well).

Dye-Loading the Cells

Fluo-3 and fluo-4 (both acetoxymethyl (AM) esters, Molecular Probes) were dissolved in low-water DMSO (purified through a molecular sieve) with 10% pluronic acid to a stock concentration of 1 mM.

The cells were dye-loaded in regular culture medium (see above) that also included fluo-3 or fluo-4 at the concentrations given in Table 1. To prevent dye extrusion by the anion exchange protein, the loading medium was supplemented with 2.5 mM probenicid, and 20 mM HEPES, pH 7.3, was added to maintain optimal pH. To compare the loading efficacy of fluo-3 and fluo-4, dye concentration and loading time were varied from those in a standard loading procedure* (2 µM fluo-3, 1.0 hour loading). After loading, a Labsystems Cellwash 4 was used to wash the cells three times with HBSS containing 20 mM HEPES and 2.5 mM Probenicid. After the last wash step, 100 µl buffer per well was left in the microplate.

The conditions defined in this investigation are not universally optimal. Optimal conditions will depend on the specific application and will be affected by many factors, e.g. cell type, receptor type, expression status of the receptor, and coupling of the receptor. Therefore, some users might need to apply higher dye concentrations to gain higher loading intensities and signals in their assays.

Fluorescent Dye Concentration Loading time
fluo-3 2 µM 0.5 hour
fluo-3* 2 µM* 1.0 hour*
fluo-4 1 µM 0.5 hour
fluo-4 1 µM 1.0 hour
fluo-4 2 µM 0.5 hour
fluo-4 2 µM 1.0 hour

Table 1: Evaluated loading conditions (* = standard condition).


Agonist Stimulation of Cells

In the FLIPR system, 50 µl aliqouts of increasing concentrations of the agonist carbachol (at 3x their final concentration) were added to the cell plate at 50 µl/sec. The final concentrations in the wells were: 2.5 µM, 1.25 µM, 625 nM, 312 nM, 156 nM, 78 nM, 39 nM, 19 nM, and 10 nM. The FLIPR system was set for 0.4 sec exposure time, F-stop of F/2 and laser power of 400 mW, and the change in fluorescence was monitored before, during, and after the carbachol addition.


Background of Non-Loaded Cells (Table 2, Column 1)

The average background of non-loaded cells was about 1,450 counts (n = 8). This is due to the autofluorescence of the plastic and the intrinsic fluorescence of the cells, which mainly comes from the flavin coenzymes FAD and FMN: Flavin-Adenine-Dinucleotide and Flavin-Mononucleotide, whose absorption/emission wavelengths are about 450/515 nm (J. Histochem Cytochem 27:44 [1979]; J. Histochem Cytochem 27:36 [1979]).

Background of Loaded Cells (Table 2, Columns 2 and 3)

The background of loaded cells varied with different loading conditions. Lengthening the loading time for either fluo-3 or fluo-4 from 0.5 hour to 1.0 hour increased the basal fluorescence intensity (indicated by an increase in basal counts). Increasing the dye concentration from 1 µM to 2 µM also increased the basal fluorescence intensity of fluo-4 (not tried with fluo-3). For the conditions tested, not surprisingly, the basal fluorescence intensity increased with increasing dye concentration and increasing loading time.

Background of Loaded Cells (Table 2, Columns 2 and 3)

The two highest concentrations of carbachol (1.25 µM and 2.5 µM) both led to a maximal stimulation of the cells as measured by calcium efflux. Therefore the average fluorescent counts from the cells stimulated with these two saturating concentrations of carbachol was taken as the maximal response.

Cells exposed to 1 µM fluo-4 for a 0.5 hour loading time gave a robust signal of 10,994 counts above basal fluorescence when stimulated. Doubling the incubation time to 1.0 hour and the concentration to 2 µM significantly increased the signal to 24,289 counts. Increasing the loading time for fluo-3 also resulted in higher signal when stimulated (0.5 hour: 6525 counts; 1.0 hour: 9765 counts).

In summary, increasing the dye concentration or loading time resulted in an increase in both the background and the stimulated signal.

Induction Level of the Signal (Table 2, Column 6)

To estimate the induction level of calcium upon stimulation, the corrected response of maximally stimulated cells (column 5) was divided by the corrected background of non-stimulated, loaded cells (column 3).

By comparing 0.5 hour loading time with 1.0 hour loading time (with a fixed dye concentration) one can see that fluo-3 (2 µM) and fluo-4 (1 µM and 2 µM) produce a higher degree of induction with a shorter incubation time (2.6→3.1 / 2.8→3.2 / 2.9→3.6).

By comparing the response to 1 µM and 2 µM fluo-4 (with a fixed loading time) it also appears that a higher dye concentration also leads to a higher degree of induction.

By comparing fluo-3 with fluo-4 under the same loading conditions (both 2 µM and 0.5 hour or 1.0 hour loading), it appears that the fluo-4 dye produces a higher magnitude of induction.

DyeConcent.Loading time(1) Unloaded Cells Background(2) loaded Cells Background(2) Unloaded Cells Background correct. for (1)(4) Stimulated Cells(5) Stimulated Cells correct. for (1)(6) Induction Level = (5)/(3)

Dye Concent. Loading time (1) Unloaded Cells Background (2) loaded Cells Background (2) Unloaded Cells Background correct. for (1) (4) Stimulated Cells (5) Stimulated Cells correct. for (1) (6) Induction Level = (5)/(3)
fluo-3 2 µM 0.5 hour 1,469
+/- 72
+/- 234
2,111 7,994
6,525 3.1
fluo-3* 2 µM* 1.0 hour* 1,469
+/- 72
+/- 310
3,689 11,234
+/- 994
9,765 2.6
fluo-4 1 µM 0.5 hour 1,415
+/- 58
+/- 422
3,442 12,409
+/- 981
10,994 3.2
fluo-4 1 µM 1.0 hour 1,458
+/- 69
+/- 485
5,897 17,721
+/- 1,360
16,263 2.8
fluo-4 2 µM 0.5 hour 1,415
+/- 58
+/- 276
4,118 16,242
+/- 830
14,827 3.6
fluo-4 2 µM 1.0 hour 1,458
+/- 69
+/- 741
8,451 25,747
+/- 1,814
24,289 2.9

Table 2: CHO-M1 loaded with fluo-3 (variable loading time) or fluo-4 (variable loading time and dye concentration).

In columns (1) to (5) fluorescent counts are given. * = Standard conditions
(1) Background from unloaded cells and plastic; n = 8
(2) Dye loaded, unstimulated cells; n = 44
(3) Dye loaded, unstimulated cells, corrected for background fluorescence
(4) Maximally (1.25 or 2.5 µM carbachol) stimulated cells; n = 8
(5) Maximally (1.25 or 2.5 µM carbachol) stimulated cells, corrected for background
(6) Induction level= (5)/(3)

Comparison of dose response curves

Figure 1 shows the dose response curves generated for the 6 different loading conditions (values expressed as a percentage of the maximum for each individual condition). As expected, the response curves are close together and neither different dyes nor different loading conditions significantly change them (and therefore the corresponding EC50 values are also unchanged, remaining between 70 and 95 nM).

Figure 1: Dose response curves for different dyes, dye concentrations, and loading times (n = 4).


These results show that fluo-4 may offer several advantages over the standardly used fluo-3. When using 2 µM fluo-4, the loading time can be decreased from 1.0 hour to 0.5 hour with equivalent or better loading than that seen with fluo-3.

Decreased loading times are beneÞcial not only because they save time, but also because the reduced exposure time may also decrease potential toxic effects of the calcium-sensitive dye. In addition, shorter loading times reduce the amount of dye that moves from the cytoplasm into intracellular compartments (e.g. mitochondria).

The decreased loading times possible when using fluo-4 also improves the signaltonoise ratio by increasing the ratio of the fluorescence produced by stimulated versus unstimulated cells. This could be because decreased dye exposure may be more favorable to cell viability. The reduced relative background fluorescence may also be attributable to a reduction in the amount of dye accumulated in the intracellular compartments.

This study has been performed using a CHO cell line, which is frequently used in HTS and in the FLIPR system. However, dyes behave differently in different cell lines and it is important that appropriate dye and loading conditions are tested for each new cell line being studied.

Learn more >>

Download PDF