Extracellular pH modifies mitochondrial control of capacitative calcium entry in Jurkat cells.

It was found that a collapse of the mitochondrial calcium buffering caused by the protonophoric uncoupler CCCP, antimycin A plus oligomycin, or the inhibitor of the mitochondrial Ca2+/Na+ exchanger led to a strong inhibition of thapsigargin-induced capacitative Ca2+ entry (CCE) into Jurkat cells suspended in a medium at pH 7.2. The effect of these inhibitors was markedly less significant at higher extracellular pH. Moreover, dysfunction of the mitochondrial calcium handling greatly decreased CCE sensitivity to extracellular Ca2+ when the pH of extracellular solution was 7.2 (apparent Kd toward extracellular Ca2+ rose from 2.3 +/- 0.6 mm in control cells to 11.0 +/- 1.7 mM in CCCP-treated cells) as compared with pH 7.8 (apparent Kd toward extracellular Ca2+ increased from 1.3 +/- 0.4 mM in control cells to 2.4 +/- 0.4 mM in uncoupler-treated cells). Changes in intracellular pH triggered by methylamine did not influence Ca2+ influx. This suggests that, in Jurkat cells, store-operated calcium channels sense extracellular pH change as a parameter that modifies their sensitivity to intracellular Ca2+. In contrast, in human osteosarcoma cells, changes in extracellular pH as well as mitochondrial uncoupling did not exert any inhibitory effects on CCE.

Opening of the plasma membrane (PM) 1 calcium channels and, in consequence, Ca 2ϩ flux into electrically non-excitable cells is triggered by a depletion of intracellular calcium stores in the lumen of the endoplasmic reticulum (ER). This regulatory mechanism leading to the PM calcium permeability is known as capacitative calcium entry (CCE) (1,2). This is the most common mechanism of Ca 2ϩ influx, at least in cells in which inositol 1,4,5-trisphosphate triggers the depletion of ER. In fact, it is universal for all types of electrically non-excitable cells and has also been postulated for some electrically excitable cells (1,3). In some cell types, e.g. in T lymphocytes, in which ER calcium stores are relatively small and thus contribute only slightly to the overall [Ca 2ϩ ] c signal, calcium influx via CCE represents the major element in the cellular calcium signaling (for review, see Ref. 4). Full activation of CCE may be obtained by the action of thapsigargin, a specific inhibitor of ER Ca 2ϩ -ATPase (5).
Ca 2ϩ influx into mitochondria is driven by the electrical potential across the inner mitochondrial membrane (6). The mitochondrial Ca 2ϩ uniporter, which is a major route of calcium entry into mitochondrial matrix, exhibits a high K d (Ͼ10 M) value for calcium ions, whereas the bulk [Ca 2ϩ ] c varies between 100 nM in resting cells and 1 M in cells responding to a stimulus. Because of the discovery that [Ca 2ϩ ] c in a close proximity of ER or of PM calcium channels may reach higher levels than measured in the bulk phase, it has been shown that, in these cellular microdomains, mitochondrial Ca 2ϩ accumulation occurs efficiently. Thereby mitochondria may play a role in the regulation of intracellular calcium signals (7,8). Respiring mitochondria located in the close vicinity of ER can buffer Ca 2ϩ released from these stores, and this may regulate the opening probability of Ca 2ϩ -sensitive ER calcium channels. Additionally, mitochondria compete with the store-loading activity of Ca 2ϩ -ATPase, counteracting the refilling of the calcium stores. This results in a more complete depletion of ER and eventually a more efficient activation of CCE (9,10). Mitochondria can take up Ca 2ϩ entering cells via plasma membrane channels. This may protect CCE from a feedback inhibition exerted by the excess of calcium accumulated in the subplasma membranous space, close to the mouths of calcium channels (7, 10 -12). Moreover, not only the buffering but also the signaling roles of mitochondria in the regulation of calcium influxes have been postulated (13,14).
The role of mitochondria in the regulation of calcium influx into electrically non-excitable cells has been intensively investigated (7, 9 -11, 15). It was found that a decrease of the mitochondrial membrane potential inhibits Ca 2ϩ flux into Jurkat cells activated by depletion of intracellular calcium stores (11,12). This effect was attributed to the feedback inhibition of CCE by Ca 2ϩ entering the cell. De-energized mitochondria could not buffer the excess of calcium and protect CCE against such inhibition.
Previously, we have documented that a decrease in the electrochemical potential of the mitochondrial inner membrane results in a decrease of the initial rate of calcium influx into Jurkat cells pretreated with thapsigargin and that this effect was strongly dependent on extracellular pH (16). We have proposed that PM calcium channels in Jurkat cells could act as pH sensors, which modify their own sensitivity to intracellular Ca 2ϩ dependently on extracellular pH.
In this study, we describe further progress in the study of mitochondrially dependent CCE regulation in electrically nonexcitable cells. We demonstrated that, in Jurkat cells, not only dissipation of the mitochondrial potential but also other disturbances in mitochondrial Ca 2ϩ handling, e.g. inhibition of the Na ϩ /Ca 2ϩ exchanger, cause a decrease in the initial rate of calcium influx in an extracellular pH-dependent manner. At pH 7.2, the CCE is depressed after mitochondrial calcium buffer blocking due to a marked increase of CCE K d for extra-cellular Ca 2ϩ . In contrast, in human osteosarcoma cells, changes in extracellular pH as well as mitochondrial uncoupling do not exert any inhibitory effects on Ca 2ϩ entry.

MATERIALS AND METHODS
Chemicals-High glucose Dulbecco's modified Eagle's medium, glutamine, and fetal bovine serum were purchased from Invitrogen. RPMI 1640 medium was from the Institute of Immunology and Experimental Therapy (Wrocław, Poland). Fura-2/AM and BCECF/AM were from Molecular Probes (Eugene, OR). Streptomycin plus penicillin for cell culture as well as thapsigargin, CCCP, oligomycin, antimycin A, and ionomycin were purchased from Sigma, and CGP 37157 was from Tocris Cookson Ltd. (Northpoint, Avonmouth, Bristol, UK). Other chemicals were of analytical grade.
Cytosolic free Ca 2ϩ was measured with Fura-2 (17). The cells were loaded with this probe in the culture medium supplemented with 1 M Fura-2/AM at 37°C for 15 min during incubation. After washing by centrifugation (Jurkat cells) in BSS supplemented with 0.1 mM CaCl 2 , the cells were suspended in nominally calcium-free BSS containing 0.05 mM EGTA, instead of CaCl 2 , and used for experiments in which [Ca 2ϩ ] c was monitored. Osteosarcoma cells were rinsed twice with the same BSS as above. The cover glasses were installed in the fluorometer cuvettes. Where indicated, the mixtures were supplemented with uncouplers or inhibitors and an appropriate amount of CaCl 2 . The fluorescence was measured at 30°C in a Shimadzu RF5000 fluorometer set in the ratio mode using 340 and 380 nm as excitation wavelengths and 510 nm as the emission wavelength. The time resolution of the measurements was 1 s. The calibration in each run was performed using externally added 3 mM CaCl 2 and 3 M ionomycin or 0.003% digitonin. Initial rates of calcium entry were expressed as a tangent of the curve representing increases in [Ca 2ϩ ] c , calculated immediately after the addition of CaCl 2 .
Both types of cells treated with CCCP were preincubated with 0.12 M oligomycin to inhibit mitochondrial ATPase and thereby to protect the cells from ATP depletion under the conditions of collapse of electric potential across the inner mitochondrial membrane by mitochondrial uncouplers. The addition of oligomycin was also required to reduce mitochondrial potential in cells preincubated with antimycin A by preventing restoration of electric potential across the inner mitochondrial membrane due to the proton-pumping function of mitochondrial ATPase. The same amount of oligomycin was added to the control samples. We have shown previously (12) that 0.12 M oligomycin does not interfere with [Ca 2ϩ ] c transients induced by the addition of thapsigargin plus CaCl 2 to Jurkat cells and that glycolytic ATP production covered the needs of the cells, even under full inhibition of oxidative phosphorylation. In cells treated with CGP 37157, no oligomycin was added.
Measurements of Intracellular pH-Cytosolic pH was measured with the fluorescent probe BCECF (18). The cells were loaded with this probe and rinsed exactly as described for Ca 2ϩ measurements with Fura-2. The fluorescence was measured at 30°C in a Shimadzu RF5000 fluorometer set in the ratio mode using 500 nm (for deprotonated BCECF) and 450 nm (for protonated BCECF) as excitation wavelengths and 530 nm as the emission wavelength. The probe was calibrated by titration with NaOH or HCl in the presence of 0.003% digitonin. The purpose of cytosolic pH measurement was to estimate changes in this parameter because of switching extracellular pH between 7.2 and 7.8. It was also necessary for adjustment of the amount of methylamine used for the selective enhancement of intracellular pH. Fig. 1 shows the effects of CCCP (Fig. 1A), antimycin A (Fig.  1B), and CGP 37157 (Fig. 1C) on calcium influx to Jurkat cells suspended in BSS at pH 7.2. The first part of each trace reflects [Ca 2ϩ ] c response to thapsigargin; transient increase in [Ca 2ϩ ] c represents calcium release from the ER followed by Ca 2ϩ extrusion to the extracellular space. Subsequent addition of CaCl 2 results in a potent increase in [Ca 2ϩ ] c because of the activation of CCE. Decrease in the mitochondrial membrane potential caused by the action of CCCP or antimycin A plus oligomycin as well as the inhibition of Ca 2ϩ efflux from mitochondria produced by the inhibitor of Na ϩ /Ca 2ϩ exchanger CGP 37157 causes a significant reduction of the initial rates of calcium influx. On the other hand, in Jurkat cells suspended in BSS at pH 7.8, the inhibition of CCE because of disturbed mitochondrial Ca 2ϩ buffering by CCCP, antimycin A plus oligomycin, or CGP 37157 was much less pronounced (Fig. 1,  D-F). The plateau of [Ca 2ϩ ] c following the addition of CaCl 2 reflects a steady-state balance between two opposite processes: Ca 2ϩ flux into the cytosol through the CCE system and Ca 2ϩ efflux from the cells mediated by PM Ca 2ϩ -ATPase and the Ca 2ϩ /Na ϩ exchanger (this latter activity in Jurkat cells is rather negligible, if any) (19). As shown in Fig. 1, D-F, the mitochondrial uncouplers or inhibitors of the respiratory chain (antimycin A) or of the Na ϩ /Ca 2ϩ exchanger (CGP 37157) have only small effects on the maximal value of [Ca 2ϩ ] c in cells suspended in BSS at pH 7.8. Decrease in the extracellular pH from 7.8 to pH 7.2 due to the addition of HCl shifts the equilibrium between influx and efflux to lower steady-state levels of [Ca 2ϩ ] c . This results most probably from both pH-dependent inhibition of CCE (20) and pH-dependent stimulation of Ca 2ϩ efflux due to increased activity of PM Ca 2ϩ -ATPase (21,22). However, a decrease of the final [Ca 2ϩ ] c related to the reduction of BSS alkalinity is more pronounced in the cells with disturbed mitochondrial calcium buffering (Fig. 1, D-F). In other words, intracellular calcium homeostasis appeared to be more sensitive to the pH of the extracellular milieu in the cells pretreated with the mitochondrial uncouplers or inhibitors than in control cells. Because there is no reason to suspect that such modifications in the mitochondrial status might activate PM Ca 2ϩ -ATPase, these findings clearly indicate that disturbances in mitochondrial Ca 2ϩ handling inhibit the activity of CCE in a pH-dependent manner.

RESULTS
Statistical evaluation of the data described above is shown in Fig. 2. Note that the absolute values of the initial rate of calcium entry into Jurkat cells suspended in BSS at pH 7.8 are significantly higher than those in cells suspended in BSS at pH 7.2.
The relatively more potent effects of CCCP in comparison with those produced by antimycin A or CGP 37157 may be partially related to the CCCP-induced dissipation of PM electrical potential, which provides the driving force for Ca 2ϩ influx. To test this possibility, the pH-dependent effect of CCCP on CCE was studied in the cells suspended in high potassium BSS (65 mM KCl and NaCl concentration lowered to 72 mM). Under such conditions, PM depolarization occurs, and this allows us to differentiate the effect of CCCP on the mitochondria from that on the PM. Despite depolarization of the PM, CCCP still much more strongly inhibited CCE in Jurkat cells suspended in BSS at pH 7.2 (from 934 Ϯ 119 to 340 Ϯ 40) than in BSS at pH 7.8 (from 1018 Ϯ 136 to 784 Ϯ 166). As expected, the absolute values of the initial rates of Ca 2ϩ entry were significantly lower because of the depolarization of PM (compare with Fig. 2).
Alkalinization of the extracellular milieu led to the alkalinization of the cytosol but to a much smaller extent due to homeostasis of intracellular parameters (16,23,24). To support the notion that changes in extracellular pH are critical for the regulation of the CCE, Jurkat cells suspended in BSS at pH 7.2 were exposed to methylamine to increase cytosolic pH without influencing extracellular pH (25). The amount of methylamine used increased cytosolic pH by 0.2-0.3 of a unit, which corresponded to cytosolic pH change after switching the extracellular pH from 7.2 to 7.8. As shown in Fig. 3, methylamine had only a small, if any, effect on the balance between Ca 2ϩ influx and efflux in the cells treated with CCCP, whereas an increase in extracellular pH from 7.2 to 7.8 due to the addition of NaOH caused a fast and significant increase in [Ca 2ϩ ] c . Similar effects were obtained for both control and antimycin-treated cells (not shown).  A (B and E). The cells treated with 10 M CGP 37157 (C and F) (as well as corresponding controls) were not preincubated with oligomycin. Thereafter, the cells were treated with 100 nM thapsigargin and exposed to 3 mM CaCl 2 . After [Ca 2ϩ ] c balance had been established, the cell suspensions were titrated with HCl to decrease extracellular pH from 7.8 to 7.2 (D-F). Each trace represents one of a few typical experiments. DMSO, Me 2 SO. position. It revealed a significant increase in the apparent K d of the CCE system for extracellular Ca 2ϩ in cells suspended in BSS at pH 7.2 pretreated with CCCP (from 2.3 Ϯ 0.6 mM for control cells to 11.0 Ϯ 1.7 mM for cells with uncoupled mitochondria). In contrast, in cells suspended in BSS at pH 7.8, CCCP had only a slight influence on the apparent K d value for extracellular Ca 2ϩ (an increase from 1.3 Ϯ 0.4 mM to 2.4 Ϯ 0.4 mM). This approach does not allow for a full interpretation of these data as kinetic parameters of calcium channel per se, because of the high complexity of the whole-cell system. However, one could conclude that respiring mitochondria are necessary to keep Jurkat cells sensitive to extracellular calcium, only if the extracellular pH is rather low.
Interestingly, thapsigargin-induced Ca 2ϩ entry into human osteosarcoma cells suspended in BSS at pH 7.2 was not inhibited in the presence of the mitochondrial uncoupler (Fig. 6).
Moreover, in these cells, CCCP caused an enhancement of the amplitude of the Ca 2ϩ transient. The increase in extracellular pH from 7.2 to 7.8 influenced neither the initial rate of Ca 2ϩ entry nor the sensitivity of CCE to the mitochondrial uncoupler (data not shown). DISCUSSION Previously, we have shown that the uncoupling of oxidative phosphorylation by CCCP in Jurkat cells strongly decreases the rate of Ca 2ϩ influx. This effect was postulated to be dependent on the extracellular pH (12,16). In this study, we present convincing evidence indicating that not only mitochondrial uncoupling but also other disturbances in mitochondrial calcium handling may reduce the activity of CCE and that these inhibitory effects are evidently dependent on extracellular pH. Moreover, we performed quantitative estimation of the effect of the mitochondrial uncoupler on the kinetic properties of the CCE system in Jurkat cells and proposed a functional model explaining the described phenomenon.
In many cell types, especially in most electrically non-excitable ones, store-operated calcium channels require energized mitochondria for buffering the excess of Ca 2ϩ and thereby for the protection of calcium influx against feedback inhibition (10 -12, 15, 16). Inhibition of Ca 2ϩ flux into Jurkat cells suspended in BSS at pH 7.2 results not only from the reduction of mitochondrial potential due to the addition of CCCP or antimycin plus oligomycin but may also be caused by inhibition of the mitochondrial Na ϩ /Ca 2ϩ exchanger. This effect indicates that mitochondrial Ca 2ϩ buffering capacity not only depends on the mitochondrial Ca 2ϩ uptake driven by membrane potential but also on mitochondrial capability for Ca 2ϩ release. This is in agreement with the data showing that mitochondrial Ca 2ϩ uptake, followed by tunneling and release of Ca 2ϩ far from PM calcium channels, is crucial for the efficient calcium buffering in close vicinity of the PM (15).
Inhibitory effects of CCCP, antimycin A plus oligomycin, and CGP 37157 on the initial rate of calcium entry into Jurkat cells were strong when the cells were suspended in BSS at pH 7.2. In cells suspended in BSS at pH 7.8, Ca 2ϩ influx was only slightly reduced. This finding confirms our previous observation that a gradual increase in extracellular pH from 7.0 to 7.8 progressively diminishes the sensitivity of CCE to the mitochondrial uncoupler (16). A small (if any) effect of intracellular alkalinization by methylamine on [Ca 2ϩ ] c in Jurkat cells suspended at pH 7.2 (see Fig. 3) indicates that the increase in intracellular pH could not explain potent effects of extracellular alkalinization on CCE activity. This is in agreement with our previous results with NH 4 Cl used for transient alkalinization of the cellular interior (16).
On the other hand, a very small but reproducible methylamine-induced increase in [Ca 2ϩ ] c in Jurkat cells might be in agreement with the observation on human platelets, in which an increase in cytosolic pH causes slight activation of calcium influx (26). Intracellular pH may, to some extent, regulate Ca 2ϩ flux into the cell; however, these effects are small, and they could not be responsible for extracellular pH-dependent changes in the rate of CCE. This is also in agreement with observations on vascular endothelial cells (20). The finding that CCE needs functioning mitochondria for full activity in cells suspended at pH 7.2 and that this dependence is less obvious at pH 7.8 is further supported by the observation that the kinetic parameters of CCE in relation to inwardly transported Ca 2ϩ are modified by extracellular pH. A large increase in the apparent K d value for extracellular Ca 2ϩ observed in the cells, suspended in BSS at pH 7.2 and exposed to CCCP in comparison to that in the cells with coupled mitochondria, suggests a competition-like relationship between extra-and intracellular calcium ions (Fig. 5). In fact, a decrease of mitochondrial potential has a negligible influence on calcium influx into Jurkat cells suspended in BSS at pH 7.2 and exposed to a very high (Ͼ6 mM) concentration of CaCl 2 (data not shown).
On the basis of the data presented in this study, we want to support our previous model of pH-dependent regulation of the channels involved in the CCE in Jurkat cells (16). According to this concept, intracellular Ca 2ϩ could efficiently bind to and inhibit the channel only if the pH of the extracellular milieu dropped below 7.4. Under such conditions, mitochondria that accumulate calcium are crucial for keeping the channel active. Under more alkaline conditions, intracellular calcium, and in consequence, the mitochondrial energy state have only a slight influence on the kinetic parameters of CCE in Jurkat cells. In other words, the calcium channel acts as an extracellular pH sensor that may switch calcium conductivity from being sensitive to the intracellular Ca 2ϩ mode (extracellular pH 7.2) to non-sensitive (extracellular pH 7.8).
Recently, it has been demonstrated that a slight acidification of the extracellular milieu leads to protonation of the glutamate residue of the TRPV5 calcium channel protein and, in consequence, to the inhibition of calcium entry into the cell (27). It has been postulated that this pH-sensing mechanism is responsible for pH-dependent conformational changes in the channel protein. In cultured vascular A7r5 cells, external acidosis decreases CCE (28), and this effect is not mimicked by intracellular acidification (for review, see Ref. 29). Similar dependen-cies between extracellular pH and Ca 2ϩ entry were found in rabbit-resident alveolar macrophages, neurons, and pancreatic cells (23,30,31). The results presented in this study not only correspond well to these data but also indicate a novel mechanism connecting both extracellular pH and mitochondrial effects on CCE in Jurkat cells.
It seems likely that extracellular pH-dependent regulation of CCE and its sensitivity to the mitochondrial energy state might be partially responsible for the abnormal behavior of lymphocytes exposed to the acidic environment and the limited availability of oxygen under some pathological conditions (33)(34)(35).
The regulatory mechanism of calcium entry into Jurkat cells (and many other electrically non-excitable cells) described in this study is based on the feedback inhibition of PM Ca 2ϩ channels by intracellular calcium and a counteracting action of mitochondria buffering the excess of Ca 2ϩ . Surprisingly, mitochondrial de-energization does not reduce the rate of calcium entry into human osteosarcoma cells. Moreover, Ca 2ϩ transients in osteosarcoma cells preincubated with thapsigargin exhibit much higher amplitude when electrical potential across the inner mitochondrial membrane collapsed (Fig. 6). This resembles, to some extent, the effects of mitochondrial uncouplers on the calcium entry into electrically excitable cells, in which mitochondrial sequestration of Ca 2ϩ entering the excited cells decreases the amplitude of Ca 2ϩ spikes and modulates the intensity of calcium signals. The gradual release of Ca 2ϩ accumulated in the mitochondrial matrix prolongs the Ca 2ϩ signal after the initial stimulus has been turned off. Reduction of the calcium buffering capacity by mitochondria (e.g. by depolarization of the mitochondrial inner membrane) results in the increased amplitude and shortened duration of the cytosolic Ca 2ϩ signals (36 -38). The comparison of the regulatory mechanisms of calcium entry into Jurkat cells and osteosarcoma cells suggests that Ca 2ϩ fluxes into these cells occur via different channel types.