The calculation of ψ was based on the Nernst equation: ψ = RT/F*ln(Ox_i/Ox_e), where R is the universal gas constant, T is the absolute temperature, F is the Faraday's constant and Ox_i and Ox_e are the internal and external concentrations of oxonol, respectively. The calibration curve was determined using different concentrations of extracellular oxonol. Since the external and internal concentrations of the dye are equal when the ψ is the same (Ox_i = Ox_e *exp ψF/RT), one can assume a new calibration curve based on Ox_i. The ratio Ox_i/Ox_e was calculated based on the acquisition of a fixed sample (ψ equal to 0 mV) using the same Ox_e for both. Afterwards the ratio value of Ox_i/Ox_e was used to calculate the ψ based on the Nernst equation 26.

We characterized the variation of ψ in Kv1.3-transfected CHO (CHO-Kv1.3) cells exposed to different concentrations of extracellular K⁺ ([K⁺]_e = 5–145 mM). Figure 1A and 1B shows the values of ψ measured by patch-clamp and the oxonol fluorescence by FACS, respectively. The values of ψ measured by either FACS or by patch-clamp were compared (fig 1C). The curves show an overlay in the range of -40 to +10 mV, indicating the reliability of ψ measurements obtained with FACS.

Our aim was also to test if this technique could discriminate between distinct ψ generate by different concentration of pharmacological blockers. CHO-Kv1.3 cells were treated with two toxins: MgTx (black bars) or iberiotoxin (IbTx; grey bars) and the ψ was measured by FACS (fig. 1D). The IbTx was chosen as a control blocker for the Kv1.3 channel, since it is the most potent and high-affinity blocker for the high-conductance calcium-activated potassium channel (BK_Ca) and it has none or low affinity for the Kv1 channels 27. Addition of MgTx depolarized CHO-Kv1.3 cells on a dose-dependent manner shifting ψ from -50.3 ± 2.5 to -3.6 ± 2.5 when the concentration of 10 nM was used. IbTx, which is a selective blocker of the BK_Ca channel 28, had no effect on the ψ of CHO-Kv1.3 cells.

In order to evaluate the ability of this method to distinguish cell populations with different ψ, we used two established cell lines, CHO and CHO-Kv1.3 29. Figure 2 shows a representative experiment, illustrating the distribution of ψ in CHO (panels A and D), CHO-Kv1.3 (panels B and E), and in a mixed population containing CHO and CHO-Kv1.3 cells (panels C and F). The peak channel of ψ in CHO-Kv1.3 cells is shifted leftwards as compared to CHO cells, and the two cell subsets can be visually distinguished when mixed in the same sample (fig. 2A–C gray lines). The mean values of ψ were -8.7 ± 2.3 mV for CHO and -41.5 ± 1.3 mV for CHO-Kv1.3 (N = 4 different experiments) which are in accordance with data published by Defarias et al. 29. Addition of 10 nM MgTX (fig. 2A–C, black lines) had no effect on the distribution of ψ in CHO cells (fig. 2A), but it depolarized CHO-Kv1.3 cells (fig. 2B) by shifting the peak channel value of ψ to -7.5 ± 1.9 mV (P 0.0001). In the presence of MgTX, CHO and CHO-Kv1.3 subsets were indistinguishable (fig. 2C). Addition of 10 nM iberiotoxin (fig. 2D–F, black line) had no effect on the distribution of ψ of either CHO (fig. 2D) or CHO-Kv1.3 (fig. 2E) and, therefore, did not affect the ψ pattern of cell subsets when they are mixed (fig. 2F).

These results indicate that it is possible to characterize the ψ of different cell populations using FACS and to evaluate the contribution of ion channels for maintenance of ψ by using specific ion channel blockers.

Human mononuclear cells from peripheral blood (PBMC) were immunostained with CD3 and CD45RO mAb and loaded with oxonol in order to evaluate the ψ in peripheral blood lymphocytes (PBL). PBL were gated according to their physical characteristics and the patterns of CD3 and CD45RO were analyzed in Figure 3A. Two subsets of T lymphocytes (CD3⁺ cells) can be identified in relation to the expression of CD45RO. Thus, CD3⁺/CD45RO^- are in the upper left gate (R2), whereas the CD3⁺/CD45RO⁺ are in the upper right gate (R3). Figure 3B shows the CD45RO⁺ subset according to the oxonol fluorescence distribution. The distribution of ψ in CD3⁺ cells did not always follow a unimodal pattern, as the example illustrated in figure 3B. In two out of six blood donors, we obtained a bimodal distribution of ψ. Separation of CD3⁺ cells according to the expression of CD45RO allowed characterization of ψ in the two cell subsets. Thus, CD3⁺/CD45RO^- cells were hyperpolarized in relation to CD3⁺/CD45RO⁺ cells (fig. 3C), the peak channel values of ψ being -58 ± 3.6 mV and -37 ± 4.1 mV, respectively (P = 0.0087).

Figure 4 shows the histograms of ψ distribution obtained with CD3⁺ cells from three different donors (A, B and C) in control (gray line) and upon treatment with either 10 nM MgTX (black line) or high [K⁺]_e (145 mM; dotted line). The three control CD3⁺ samples represent the distinct distribution patterns of ψ. Thus, donors A and C showed a unimodal pattern on the control trace, whereas donor B showed a bimodal pattern. Addition of high [K⁺]_e depolarized all the samples to approximately 0 mV (peak channel evaluation) and generated a unimodal pattern in all histograms studied. In contrast, addition of MgTX produced diverse degrees of response, causing partial depolarization, which broadened the ψ distribution (Fig. 4A) or generated a bimodal pattern (Fig. 4C). Percentage of cells which depolarized upon treatment with MgTx and overlapped with the histogram acquire after treatment with high [K⁺]_e were determined. Addition of MgTx did not depolarize all population when compared with the high [K⁺]_e treatment; it rather depolarized a small part (black marker) which accounts for 27, 39 and 12% of CD3⁺ cells from the different blood donors (fig. 4A,B and 4C, respectively).

In order to characterize the distinct degrees of depolarization upon exposure to MgTX in activated T lymphocytes, we investigated the presence of CD45RO^- and CD45RO⁺ cells in the different samples and evaluated the sensitivity of each cell subset to MgTX. The proportion of CD45RO^- and CD45RO⁺ varied among donors and the individuals with higher amounts of CD45RO⁺ cells among T lymphocytes (N = 4) showed significant depolarization upon treatment with MgTX (from -45 to -15 mV, P = 0.0025), whereas those with low proportion of CD45RO⁺ cells (N = 2) were not affected by MgTX exposure (data not shown).

In view of these results, we analyzed the distribution of ψ and the effects of MgTX in CD45RO^- and CD45RO⁺ subsets. Figure 5 shows the results obtained in three out of six donors (panels A, B and C), which illustrate the different patterns and degrees of response. The CD45RO^- and CD45RO⁺ subsets from each donor are shown in left and right panels, respectively. The CD45RO⁺ subsets presented a unimodal distribution of ψ in all donors studied. In contrast, CD45RO^- subsets had a more variable distribution of ψ, with a bimodal pattern being seen in two out of six donors (panel B-left shows an example). Addition of high [K⁺]_e (dotted line) depolarized all cell subsets. In contrast, addition of MgTX had no significant effect in CD45RO^- subsets (P = 0.15, N = 6) but depolarized CD45RO⁺ cells, shifting the peak channel value of ψ from -37.2 ± 4.1 mV to -26.7 ± 5.1 mV (P = 0.0025, N = 6). By comparing CD45RO^- and CD45RO⁺ subsets that depolarized at the same extent as high [K⁺]_e treatment (marker), we had always 2-fold increase of percentage of cells in the later subset for each donor. Percentage of cells was 5 and 13%, 29 and 64% and 17 and 37% (CD45RO^- and CD45RO⁺, respectively; fig. 5A,B and 5C).

CHO cells were obtained from Rio de Janeiro Cell Bank (PABCAM, Federal University, Rio de Janeiro, RJ, Brazil) and CHO-Kv1.3 cells 29 were a kind gift from Dr. Maria L. Garcia (Merck & Co., Rahway, NJ, USA). Both cell lines were maintained in α-Minimum Essential Medium (Gibco-BRL – Life Technologies, Inc., Grand Island, NY, USA) supplemented with 10% (v/v) heat-inactivated bovine fetal serum (Gibco), 60 mg/L penicillin (Sigma Chemical Co., St Louis, MO, USA), 100 mg/L streptomycin (Sigma) 2.4 g/L HEPES (Sigma). Cultures were grown in a humidified incubator at 37°C, 5% CO₂.

PBMC were isolated from the peripheral blood of healthy donors by centrifugation on a Ficoll gradient. PBMC were washed and incubated with RPMI 1640 medium (Sigma) supplemented with 10% (v/v) heat-inactivated bovine fetal serum (Gibco), 60 mg/L penicillin (Sigma), 100 mg/L streptomycin (Sigma) for 30 minutes (37°C, 5% CO₂ humidified atmosphere). The study was evaluated by the National Cancer Institute (INCa-Brazil/RJ) Ethical Committee and the informed consent of all participating subjects was obtained.

Samples were fixed with ice-cold 2% formaldehyde and kept at 40°C for 60 minutes. Cells were washed with PBS and kept at room temperature before the measurements.

Bis(1,3-dibutylbarbituric acid(5)) trimethine oxonol (diBA-C4-(3)), obtained from Molecular Probes (Invitrogen, Carlsbad, CA, USA), was dissolved in DMSO and stored in aliquots (1 mM) at -20°C, under protection from light. Aliquots were added to the cell suspensions to yield the desired final concentration (5–1500 nM). The oxonol concentration of 100 nM was used for the membrane potential measurements. The DMSO final concentration (0.1% (v/v)) had no effect on cell viability. Stock solutions of MgTX (kindly provided by Dr. Maria L. Garcia, Merck & Co.) and IbTX (Alamone Labs, Jerusalem, Israel) were prepared in a saline solution with 100 mM NaCl, 20 mM Tris-HCl (pH 7.4), and 0.1% (w/v) BSA. All other reagents were of analytical grade.

Membrane potentials of CHO-Kv1.3 cells were measured by whole-cell patch-clamp recordings obtained with an EPC-7 amplifier (Axon Instruments, Foster City, CA, USA) in current clamp mode with a 2 KHz analogical filter. The signals were digitized at 5 KHz (interface DigiData 1200 – Axon Instruments) and analyzed with the software pClamp 6.0 (Axon Instruments). Membrane voltages were corrected for liquid junction potentials before patching the cell at the voltage clamp mode. Pipettes with resistances of 3–6 M, were pulled from thin-walled borosilicate glass capillaries (Rochester Scientific Co. Inc., Rochester, NY, USA), and filled with a solution containing (in mM): 140 KCl, 2 MgCl2, 0.123 CaCl2, 0.2 K2-EGTA, 10 Hepes (pH 7.4). The extracellular medium, designated PSS (physiological saline solution), contained (in mM): 140 NaCl, 5 KCl, 2 MgCl2, 1 CaCl2, 10 HEPES (pH 7.4). High K⁺ solutions (10 – 145 mM) were obtained by isotonic replacement of NaCl with KCl in the PSS. Experiments were performed at room temperature (22–25°C). The liquid junction potential correction was performed using the software JPCalc 41.

We used the method described by Krasznai et al. 26 for determination of ψ. The phosphate buffer solution (PBS) was replaced by the PSS, which is the standard solution in all experiments of electrophysiology in our laboratory. The method was validated using the PSS in CHO cells and in human lymphocytes.

PBMC were labeled with a PercP-conjugated anti-human CD3 or PE-conjugated CD45RO mouse antibody (Pharmigen, San Diego, CA, USA), Fc receptor being blocked with normal mouse serum (1:50) in PBS buffer. After a 20 minute-incubation with the antibody at 4°C, cells were washed with PSS kept at room temperature and re-suspended at a concentration of 10⁶ ml-1.

Measurements were carried out at room temperature, using a Becton Dickison FACScan flow cytometer and data were analyzed using Cell Quest or FlowJo program. Forward-scatter (FSC) and side-scatter (SSC) lights were used for gating of data acquisition. Non-viable cells were identified with propidium iodide (Sigma), and were excluded from analysis. All samples were excited with the 488 nm line and oxonol, PE and PercP fluorescence emission were captured at 530/30 nm, 585/42 nm and 670 nm long pass, respectively. The calculated values of ψ within a cell subset are presented in histogram distributions and the peak channels were used for comparative analysis. All the experiments were performed at least four times and the data are presented as mean ± standard error.

Unpaired t test was performed for comparison of the values of ψ between cell subsets and paired t test was used for comparisons of ψ after different treatments within a given cell subset. The software GraphPad Prism version 4 was applied for the analysis.