Calcium is involved in numerous physiological and pathophysiological signaling pathways. Live cell imaging requires specialized equipment and can be time consuming. A quick, simple method using a flow cytometer to determine relative changes in cytosolic calcium in adherent epithelial cells brought into suspension was optimized.
A variety of cellular processes, both physiological and pathophysiological, require or are governed by calcium, including exocytosis, mitochondrial function, cell death, cell metabolism and cell migration to name but a few. Cytosolic calcium is normally maintained at low nanomolar concentrations; rather it is found in high micromolar to millimolar concentrations in the endoplasmic reticulum, mitochondrial matrix and the extracellular compartment. Upon stimulation, a transient increase in cytosolic calcium serves to signal downstream events. Detecting changes in cytosolic calcium is normally performed using a live cell imaging set up with calcium binding dyes that exhibit either an increase in fluorescence intensity or a shift in the emission wavelength upon calcium binding. However, a live cell imaging set up is not freely accessible to all researchers. Alternative detection methods have been optimized for immunological cells with flow cytometry and for non-immunological adherent cells with a fluorescence microplate reader. Here, we describe an optimized, simple method for detecting changes in epithelial cells with flow cytometry using a single wavelength calcium binding dye. Adherent renal proximal tubule epithelial cells, which are normally difficult to load with dyes, were loaded with a fluorescent cell permeable calcium binding dye in the presence of probenecid, brought into suspension and calcium signals were monitored before and after addition of thapsigargin, tunicamycin and ionomycin.
Calcium is an important second messenger responsible for the transmission of cellular physiological and pathophysiological signals1. Cytosolic calcium concentrations are kept low through the activity of calcium pumps and calcium exchangers. The sarcoplasmic/endoplasmic reticulum calcium ATPase (SERCA) refills the endoplasmic reticulum (ER) calcium store as part of a “pump leak” system whereas the plasma membrane calcium ATPase (PMCA) extrudes calcium to the extracellular compartment, both in ATP-dependent manners2. Physiological messengers, such as hormones or neurotransmitters, transmit their signals through G-protein coupled receptors, activation of phospholipase C, which hydrolyses phosphatidylinositol 4,5-bisphosphate (PIP2) in the plasma membrane to generate diacylglycerol and inositol triphosphate (IP3)3. While diacylglycerol remains in the plasma membrane, IP3 diffuses into the cytosol and binds to IP3 receptors (IP3Rs), which are ligand activated calcium channels, found in the ER membrane and calcium is released from the ER luminal store culminating in an increase in cytosolic calcium concentrations4. An alternative route for calcium to reach the cytosol is through calcium channels and exchangers present in the plasma membrane. Crosstalk between these two compartments have been described: calcium induced calcium release (CICR) where extracellular calcium induces calcium release from ER stores5, and store operated calcium entry (SOCE) where emptying of the ER stores is sensed by STIM and causes opening of Orai calcium channels in the plasma membrane to promote refilling of the ER stores6.
Under pathophysiological conditions, the cellular calcium response is deregulated and cytosolic calcium increases in the absence of physiological stimulators1. The calcium response may be affected in a number of ways: prolonged calcium signal, delayed calcium removal from the cytosol, depletion of calcium stores or localized changes in calcium. Furthermore, the mitochondria take on a central role in buffering and releasing calcium7. Prolonged and/or supramaximal cytosolic calcium increase leads to cell death. In fact, calcium is more often than not involved in the cellular pathophysiological response and is a key event in a wide array of diseases, such as neurodegenerative disease, heart disease, cancer, bone disease and kidney disease, which are mainly associated with cell death and loss or alteration of organ or tissue function8-10. Moreover, perturbation of calcium signals has been linked to cellular adaptation and changes in cellular functions and responses.
Traditionally, calcium signals are measured with a negatively charged fluorescent calcium indicator that is loaded into cells in a cell permeable acetoxymethyl (AM) ester form. Once cleaved by cellular esterases, the fluorescent indicators remain within the intracellular compartment and increase in fluorescence intensity when calcium is bound. The most well known and used calcium indicator is the ratiometric Fura-2 developed by Roger Tsien and colleagues11. Calcium indicators with differing affinities for calcium allow different calcium pools to be monitored. Detection methods include live cell imaging, fluorescence microplate reader and flow cytometry. The relatively fast kinetics of calcium signals (usually within a few seconds to minutes) makes live cell imaging the best method for gaining the most information about characteristics of the calcium signal. Aside from the fast kinetics of calcium signals (within milliseconds), live cell imaging is a better method for studying cellular compartmentalization of calcium signaling within a single cell 12. Absolute calcium concentrations can be calculated by determining the minimum and maximum fluorescence of the calcium indicator by adding a calcium chelator and permeabilizing the cells, respectively, as described by Grynkiewicz et al.11.
The use of flow cytometry to measure calcium signals was first developed in the 1980s in immunological cells where the opportunity to measure multiple parameters, such as membrane integrity and separation of cell population, and the requirement of cells in suspension combined with the development of single wavelength calcium indicators made flow cytometry an ideal and convenientdetection method13-15. In adherent cells, live cell imaging provides the most information about calcium signaling, most importantly the kinetics, but requires a complicated setup comprising a fluorescence microscope, a perfusion system, maintenance of the cellular environment, such as temperature, and specialized microscope software for acquiring and analyzing data. Alternative methods such as a fluorescence microplate reader or pharmacological means through use of calcium chelators are incomparable in terms of gained information. Flow cytometry is becoming more widely used to monitor cytosolic calcium in adherent cells though, in the majority of studies, only end-point measurements are performed. Using the method initially developed by Gergely et al. in immunological B cells16, changes in cytosolic free calcium concentration by flow cytometry with real-time kinetics and monitoring the fluorescence intensity in individual cells from a sample population have been successfully measured in cancer epithelial cells and cancer stem cell hybrids17. This method was further adapted for use in adherent epithelial cells, which are difficult to load with calcium indicators 18.
Here, using an immortalized adherent epithelial cell line derived from the S1 segment of the rat kidney proximal tubule (WKPT-0293 Cl.2)19, we describe an optimized simplified method for measurement of changes in cytosolic calcium concentration by flow cytometry. Since many epithelial cells possess a number of organic transporters for anionic compounds, loading of cells with a calcium indicator can prove to be a challenge. To prevent the efflux of calcium indicators, probenecid, the prototypical inhibitor of organic anion transporters and originally developed to decrease renal excretion of penicillin20, is applied simultaneously in the incubation medium. The cells are maintained in monolayers for calcium indicator loading, brought into suspension and calcium signals can be detected immediately after compound addition.
1. Preparation of Reagents and Solutions
2. Cell Culture
3. Calcium Dye Loading
4. Preparation of Cells for Flow Cytometry
5. Flow Cytometer Setup
6. Flow Cytometry Measurement
7. Data Analysis
To increase cytosolic calcium, two pharmacological compounds were applied to renal proximal tubule cells (WKPT-0293 Cl.2) loaded with cell permeable calcium binding dye, Fluo-3-AM. Nontreated control samples were loaded with calcium binding dye in the presence of probenecid and were mixed but without the addition of any compounds. Thapsigargin (TG) is a classical inhibitor of the SERCA pump leading to leakage out of the ER and resulting in increased cytosolic calcium. Tunicamycin (TUN) blocks N-glycosylation of proteins leading to activation of the unfolded protein response but also increases cytosolic calcium18,21. As a positive control, ionomycin (IONO), a calcium ionophore making the plasma membrane permeable for extracellular calcium, was used. For analysis, the cell population was gated by selecting a region in the SSC/FSC dot plot (R1 in Figure 1). A density plot for fluorescence intensity vs. time was created for each data file. By using the quadrant function, the dot plot was separated into four areas representing Fluo-3 fluorescence intensity before and after compound addition and above and below the baseline fluorescence. Quantification as percentage of gated cell population in each quadrant allows determination of increased cytosolic calcium concentration. In this example, 3 µM TG causes an increase in the percentage of cells in the upper right quadrant indicating that a calcium signal has been induced. The percentage of cells increases from 42.8% in control to 51.3% (1.20 fold) in TG treated cells. Similarly, 6 µM TUN augments the percentage of cells with increased calcium fluorescence intensity to 49.9% (1.17 fold). Application of 10 µM IONO resulted in 65.5% (1.53 fold) of cells in the upper right quadrant (Figure 2). The histogram analysis resulted in 1.41 fold, 1.36 fold and 2.11 fold by TG, TUN and IONO, respectively (Figure 3) (for statistics, see 18). Finally, using the multiple region analysis mode, TG, TUN and IONO caused 1.55 fold, 1.29 fold and 4.54 fold increases in calcium signal over control, respectively (Figure 4).
The most sensitive analysis mode was the multiple region analysis. In contrast to the quadrant and histogram analysis modes, where the distribution of the cell population is quantified, the multiple region analysis mode quantifies the change in fluorescent signal from the total cell population. The largest discrepancy in the results from different analysis modes was for ionomycin that ranged from 1.53 fold increase over control from the quadrant analysis to 4.54 fold from the multiple region analysis. The compounds tested did not exhibit such variance. However, throughout all analysis modes, the extent of calcium signal evoked remained the same though the absolute values differed, i.e., IONO had the largest effect followed by TG and then TUN.
Application of TG, TUN or IONO results in permanent changes in cytosolic calcium. To determine whether physiological calcium signals, which are transient, can be measured using this method, ATP was applied to WKPT-0293 Cl.2 cells. ATP evoked a rapid and transient calcium signal, which could not be recorded in its entirety by the flow cytometry method. ATP concentrations at 20 – 100 µM were difficult to detect therefore the concentration was reduced to “slow down” the calcium signals. As shown in Figure 5, a change in cytosolic calcium could be measured after addition of 5 µM ATP. Still, the peak of fluorescence change was not recorded. Using the quadrant analysis (Figure 5B) and histogram analysis (Figure 5C) modes, no difference in calcium signals in ATP could be detected due to the transient nature of the calcium signal. However, with selected parts of the trace, the multiple rectangular region analysis mode recorded a 1.85 fold increase in calcium signal immediately after resumed measurement (Figure 5A).
Figure 1. SSC versus FSC dot plot. The integrity of the cell population is assessed using side scatter (SSC) and forward scatter (FSC). A region is selected (R1) for further analysis. Dead cells and cell debris, which have reduced light scattering, are excluded. A representative dot plot is shown.
Figure 2. Quadrant analysis. Baseline fluorescence of calcium binding dye-loaded WKPT-0293 Cl.2 cells was measured for 50 sec. Measurement was stopped, compound or diluent was added, vortexed to mix and measurement was resumed immediately for a total of 204.8 sec. Quantification was performed using quadrant analysis. Please click here to view a larger version of this figure.
Figure 3. Histogram analysis. Histogram plot of calcium binding dye-loaded WKPT-0293 Cl.2 cells treated with TG, TUN or IONO. Using the marker function, the percentage of gated cells with high calcium binding dye fluorescence intensity was quantified by positioning the starting point of the marker at the peak of the control curve and ending at the maximum fluorescence. Please click here to view a larger version of this figure.
Figure 4. Multiple region analysis. Baseline fluorescence of calcium binding dye-loaded WKPT-0293 Cl.2 cells was measured for 50 sec. Measurement was stopped, compound or diluent was added, vortexed to mix and measurement was resumed immediately for a total of 204.8 sec. Quantification was performed using multiple rectangular region analysis. Please click here to view a larger version of this figure.
Figure 5. Analysis of ATP-induced calcium signals. Baseline fluorescence of calcium binding dye-loaded WKPT-0293 Cl.2 cells was measured for 50 sec. Measurement was stopped, 5 µM ATP was added, vortexed to mix and measurement was resumed immediately for a total of 204.8 sec. Data was quantified using multiple rectangular region analysis (A), quadrant analysis (B) or histogram analysis (C). For the multiple region analysis, only the transient calcium signal was analyzed. Representative data or analysis from n = 3 – 7 are shown. Please click here to view a larger version of this figure.
Calcium is a key event in multiple cellular processes acting as a second messenger signaling molecule and also as a means to communicate between the ER and mitochondria. The ability to measure changes in free cytosolic calcium concentrations is a useful technique that can apply to all areas of cell biology. There are various methods to detect cytosolic calcium signals. Classically, signals from a ratiometric calcium indicator, such as Fura-2, are monitored in live cells using a fluorescence imaging setup and absolute calcium concentrations can be derived from determining minimum and maximum fluorescence intensities. However, this technique requires specialized equipment, such as a fluorescence microscope, a live cell imaging set up and analysis software. Here, we describe a simple method for calcium detection in the cytosol of adherent epithelial cells using flow cytometry, which is accessible to most laboratories.
Epithelia are in possession of an array of transporters responsible for the transport of metabolites, xenobiotics and drugs. Of particular importance is the SLC22A superfamily, to which organic cation and organic anion transporters belong 22, and the ATP-binding cassette (ATP) superfamily, which counts the multidrug resistance P-glycoprotein ABCB1 and multidrug resistance proteins (MRPs) amongst its members 23. Initial experiments using the rat kidney proximal tubule cell line WKPT-0293 Cl.2 were performed as described in the paper by Gergely et al.16. The cells were brought into suspension and loaded with Fluo-3 AM. However, only a modest change was observed with 10 µM ionomycin (38.14% versus 50.95% in control by quadrant analysis or 1.78 fold by multiple region analysis). We hypothesized that epithelial transporters were responsible for the poor loading of the cells with Fluo-3 AM. To circumvent this problem, probenecid and PSC833, inhibitors of organic anion transporters and the multidrug resistance P-glycoprotein, respectively, were applied simultaneously with the calcium binding dye to cell monolayers. Probenecid improved calcium binding dye loading, as indicated by an increase in baseline fluorescence intensity, whereas PSC833 had no effect, which is in contrast to a previous report in Chinese hamster ovary cells24. This difference probably lies in the levels of ABCB1 in the different cell lines; the WKPT-0293 Cl.2 cell line harbors a low endogenous level of ABCB1, which can be induced by toxic substances, such as toxic metals25,26. The use of probenecid to improve calcium dye loading can be extended to other cell types that exhibit probenecid-sensitive transport systems, such as cells of the nervous system, blood-brain barrier and the liver, where calcium signaling may play a role in intracellular signaling pathways.
The easy detection of calcium binding to Fluo-3 lies in its requirement of only one excitation wavelength (506 nm) and one emission wavelength (525 nm). However, the data generated with single wavelength calcium binding dyes, as a non-ratiometric dye, can only be used to quantify relative increases in cytosolic calcium over non-treated controls. To determine absolute calcium concentrations, a ratiometric dye is necessary because a spectral shift occurs after calcium binding that allows for correction of unequal dye loading and bleaching. The ratiometric dye Indo-1 has been used in detection of calcium mobilization in immunological cells by flow cytometry27,28 but so far, this has not been performed in epithelial cells. Using the two emission spectra of UV-excitable Indo-1 (400 nm when calcium bound and 475 nm when calcium free), absolute cytosolic calcium concentrations can be determined as described by Grynkiewicz et al.11.
Our example data employed pharmacological compounds that induce a permanent change in cytosolic calcium. Physiological calcium signals are of low intensity, display fast kinetics and are transient, thus the question remains as to whether the flow cytometry method is sensitive enough to detect physiological calcium signals. Experiments with ATP were performed with the WKPT-0293 Cl.2 cell line, which harbors purinergic receptors as demonstrated by increased calcium signal upon application of ATP (measured by live cell imaging). Using the flow cytometer, increased cytosolic calcium could be observed with 1 – 100 µM ATP, but the rapid nature of the signal combined with the delay in resuming measurement after compound addition meant that only a portion of the signal could be recorded. Thus a major disadvantage of the flow cytometry method is its limited capacity to record rapid (<10 sec) and transient calcium signals; rather the method is more suitable for stimuli that give rise to slower transient signals (>10 sec) or persistent elevated calcium signals. These limitations could be circumvented by reducing stimulator concentration or by applying inhibitors for calcium reuptake mechanisms to retain the released calcium in the cytosol but caution must be taken to select an inhibitor concentration that does not increase cytosolic calcium per se.
The information acquired by live cell imaging and flow cytometry are quite different. With live cell imaging, the high resolution means that spatial changes in calcium can be measured by selecting intracellular regions of single cells for imaging. Furthermore, fluorescence signals are acquired nonstop every 1 – 2 sec therefore rapid calcium signals can be recorded. In contrast, flow cytometry is more suited for slower and long lasting calcium signals and measures calcium signals from a whole cell population. Only relative changes in calcium signals can be quantified with the flow cytometry method. But, the flow cytometry method has a number of advantages over conventional live cell imaging: 1) high throughput screening of compounds (stimulators and inhibitors of calcium signaling) can be performed; 2) specific imaging expertise is not required; 3) research technicians can perform the measurements; 4) clinical and non-clinical laboratories that do not have the means to perform high resolution imaging and ratiometric calcium measurements can investigate calcium signaling. Taken together, the flow cytometry method is useful if the investigator initially wishes to establish a role for calcium and could easily apply inhibitors to determine upstream and downstream signaling pathways, whereas live cell imaging would be required to characterize the calcium signal (e.g., kinetics, intensity, spatial changes) in further detail.
In summary, we describe a quick, simple and sensitive method for the detection of real-time relative changes in cytosolic calcium in difficult-to-load adherent epithelial kidney cells, which have been brought into suspension. For experimenters without access to a flow cytometry, this method can be easily applied to a spectrofluorimeter with the cell suspension in a cuvette.
The authors have nothing to disclose.
Research in the laboratories is funded by the Fritz-Bender Foundation, Munich, Germany (to T.D.) and a University of Witten/Herdecke internal research grant (to W.-K.L). We would like to thank Prof Dr. Dr. Frank Thévenod (Institute for Physiology, Pathophysiology & Toxicology, University of Witten/Herdecke) for helpful suggestions.
Name of Reagent/ Equipment | Company | Catalog Number | Comments/Description |
Fluo-3, AM | Invitrogen | F-1241 | Cell permeable |
Probenecid | Sigma-Aldrich | P-8761 | Water insoluble, dissolve in 1 N NaOH |
0.05% Trypsin-EDTA (1X) | Invitrogen | 25300-062 | |
Ionomycin | Sigma-Aldrich | I9657 | |
Thapsigargin | Tocris Bioscience | 1138 | |
Tunicamycin | Sigma-Aldrich | T7765 | |
FACSCalibur Flow Cytometer + CELLQuest software | Becton Dickinson | ||
Windows Multiple Document Interface for Flow Cytometry (WinMDI) | Joe Trotter, The Scripps Institute | Please note that this is an older 16-bit application that reads FCS 2.0 compliant files but does not recognize FCS 3.0 digital data. | |
WKPT-0293 Cl.2 rat kidney proximal tubule cell line | Made available by Dr. Ulrich Hopfer (Department of Physiology & Biophysics, Case Western Reserve University, Cleveland, OH) |