A novel regulatory mechanism of the ... - The FASEB Journal

The FASEB Journal express article 10.1096/fj.02-0553fje. Published online October 4, 2002.

A novel regulatory mechanism of the mitochondrial Ca2+ uniporter revealed by the p38 mitogen-activated protein kinase inhibitor sb202190 Mayte Montero, Carmen D. Lobatón, Alfredo Moreno, and Javier Alvarez Instituto de Biología y Genética Molecular (IBGM), Departamento de Bioquímica y Biología Molecular y Fisiología, Facultad de Medicina, Universidad de Valladolid and Consejo Superior de Investigaciones Científicas (CSIC), Ramón y Cajal, 7, E-47005 Valladolid, SPAIN. Corresponding author: Javier Alvarez, Instituto de Biología y Genética Molecular (IBGM), Departamento de Bioquímica y Biol. Mol. y Fisiología, Facultad de Medicina, Ramón y Cajal, 7, E-47005 Valladolid, Spain. E-mail: [email protected] ABSTRACT It is widely acknowledged that mitochondrial Ca2+ uptake modulates the cytosolic [Ca2+] ([Ca2+]c) acting as a transient Ca2+ buffer. In addition, mitochondrial [Ca2+] ([Ca2+]M) regulates the rate of respiration and may trigger opening of the permeability transition pore and start apoptosis. However, no mechanism for the physiological regulation of mitochondrial Ca2+ uptake has been described. We show here that SB202190, an inhibitor of p38 mitogen-activated protein (MAP) kinase, strongly stimulates ruthenium red-sensitive mitochondrial Ca2+ uptake, both in intact and in permeabilized HeLa cells. The [Ca2+]M peak induced by agonists was increased about fourfold in the presence of the inhibitor, with a concomitant reduction in the [Ca2+]c peak. The stimulation occurred fast and was rapidly reversible. In addition, experiments in permeabilized cells perfused with controlled [Ca2+] showed that SB202190 stimulated mitochondrial Ca2+ uptake by more than 10-fold, but only in the physiological [Ca2+]c range (1–4 µM). Other structurally related p38 MAP kinase inhibitors (SB203580, PD169316, or SB220025) produced little or no effect. Our data suggest that in HeLa cells, a protein kinase sensitive to SB202190 tonically inhibits the mitochondrial Ca2+ uniporter. This novel regulatory mechanism may be of paramount importance to modulate mitochondrial Ca2+ uptake under different physiopathological conditions. Key words: calcium • mitochondria • aequorin • MAP kinase

I

n recent years, it has becoming increasingly clear that not only does Ca2+ uptake by mitochondria serve to increase the mitochondrial Ca2+ levels ([Ca2+]M) and stimulate respiration after cell stimulation, but it also has large impact on the regulation of the overall cell Ca2+ homeostasis (1–3). A large fraction of the Ca2+ entering the cell during stimulation appears to be transiently buffered by mitochondria (4–6), which can suffer reversible and repetitive near millimolar [Ca2+]M transients (7). Therefore, mitochondria may become important regulators of the cytosolic [Ca2+] ([Ca2+]c) and thus of many different [Ca2+]c-dependent

processes. In addition, under some conditions, mitochondrial Ca2+ loading may lead to opening of the permeability transition pore and trigger the apoptotic pathway (1). However, in spite of the importance of mitochondrial Ca2+ uptake for many different aspects of cell function, little is known about its physiological regulation. Ca2+ enters mitochondria through the so-called Ca2+ uniporter–a transport system whose molecular substrate is still unknown but that behaves functionally as a specific Ca2+ channel activated by cytosolic Ca2+ in the micromolar range (2, 8, 9). The presence of a different mechanism, called rapid uptake mode, adapted to sequester Ca2+ from trains of physiological Ca2+ transients has been also suggested (8), although no direct evidence for this mechanism has been obtained in intact cells. The Ca2+ uniporter is inhibited by ruthenium red and suffers a complex modulation by aliphatic polyamines such as spermin or aminoglucosides (9, 10), but no physiological mechanisms of regulation of the uniporter have been described. We describe here for the first time a regulatory mechanism that is able to increase the rate of mitochondrial Ca2+ uptake by more than one order of magnitude in a few seconds and operates only at low, physiological [Ca2+]c. MATERIALS AND METHODS Cell transfection and generation of stable clones The construction strategy of the mutated mitochondrially targeted aequorin chimera (mitmutAEQ) has been described previously (7). HeLa cells were grown in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal calf serum. For generating HeLa cell clones stably expressing mitmutAEQ, two 10-cm dishes of HeLa cells containing 1.5×106 cells each were transfected with 8 µg/dish of mitmutAEQ cloned into the pcDNA3.1 plasmid. Selection was carried out with 0.8 mg/ml G418, and 42 clones were isolated and tested for aequorin expression by measuring the light emission of coelenterazine-reconstituted cell plates. Clone MM5 was the highest producer and was used in the experiments presented here. Similar data were obtained using other clones or wild-type cells transiently transfected with the mitmutAEQ/pcDNA3.1 plasmid. Cytosolic aequorin cDNA was obtained from Molecular Probes (Eugene, OR) and cloned into the pcDNA3.1. plasmid. Transfections were carried out using Fugene. [Ca2+]M and [Ca2+]c measurements Cell clones were plated onto 13-mm round coverslips. For aequorin reconstitution, HeLa cells expressing cytosolic aequorin were incubated for 1–2 h at room temperature with 1 µM of wildtype coelenterazine, and cells expressing mitmutAEQ were incubated for 1–2 h at room temperature with 1 µM of either wild-type coelenterazine or coelenterazine n, in standard medium containing 145 mM NaCl, 5 mM KCl, 1 mM MgCl2, 1 mM CaCl2, 10 mM glucose, and 10 mM HEPES, pH 7.4. Cells were then placed in the perfusion chamber of a purpose-built luminometer thermostatized at 37ºC. For experiments with permeabilized cells, cells expressing mitmutAEQ reconstituted with coelenterazine n were placed in the luminometer. Then, standard medium containing 0.5 mM EGTA instead of Ca2+ was perfused for 1 min, followed by 1 min of intracellular medium (130 mM KCl, 10 mM NaCl, 1 mM MgCl2, 1 mM K3PO4, 0.5 mM EGTA, 1 mM ATP, 20 µM ADP, 2 mM succinate, 20 mM Hepes, pH 7) containing 100 µM digitonin. Then, intracellular medium without digitonin was perfused for 1 min, followed by buffers of

known [Ca2+] between 1.5 and 5.5 µM, prepared in intracellular medium, using HEDTA/Ca2+/Mg2+ mixtures. To calibrate the data obtained in terms of [Ca2+], for each experiment, we needed to know the total amount of luminescence that could be emitted by the sample. For that, at the end of every experiment, we perfused lysis solution containing detergent (100 µM digitonin) and excess Ca2+ (10 mM) to measure all the remaining aequorin luminescence. To transform luminescence data in [Ca2+], a computer program subtracted the background and calculated the fractions L/Lmax at every point along the experiment. L is the luminescence value at every point (minus the background), and Lmax is the integral of luminescence (minus the background) from that point to the end of the experiment. L/Lmax is therefore the fraction of the total remaining luminescence recorded from the sample in 1 s at each point. For every [Ca2+], a particular fraction of the total aequorin luminescence is emitted per second, and this fraction increases rapidly with the [Ca2+]. L/Lmax values were then transformed into [Ca2+] values by using the following mathematical algorithm:

[Ca ] 2+

1/ n

(in M ) = ratio + (ratio • KTR ) − 1 , where ratio =  L  KR − (ratio • KR )  Lmax • λ 

This algorithm and the parameters KR, KTR, and n derive from a mathematical model proposed originally to explain from a molecular point of view the Ca2+ dependence of aequorin luminescence (11) but can be used as a simple mathematical transformation independent of the model. The parameter λ is the rate constant for aequorin consumption at saturating [Ca2+]. This parameter was not included in the original description of the algorithm (12), because the maximum rate constant of native aequorin reconstituted with wild-type coelenterazine is 1.0 s–1. Reconstitution with coelenterazine n reduces considerably the maximum rate constant, and this allows recording high [Ca2+] values with slower aequorin consumption. The values of the parameters of the algorithm for the different aequorin-coelenterazine combinations used here have been obtained previously (7, 13, 14). For native aequorin reconstituted with wild-type coelenterazine at 37ºC, KR = 4.81×107; KTR = 601; n = 2.3 ; λ = 1. For mutated aequorin reconstituted with wild type coelenterazine at 37ºC, KR = 1.61×107; KTR = 2.2×104; n = 1.43; λ = 1. For mutated aequorin reconstituted with coelenterazine n at 37ºC, KR = 8.47×107; KTR = 1.656×105; n = 1.2038 ; λ = 0.138. Emission of light by aequorin is irreversible in our conditions, and this is usually referred to as “aequorin consumption.” Calculation of the percent of aequorin consumed during a particular stimulation is useful to evaluate whether the kinetics of the [Ca2+] peak may be somehow affected by saturation of the probe. For these calculations, the total light emitted along the experiment (including that emitted after cell lysis) was taken as 100%. Statistical values are given as mean ±SE. Wild-type coelenterazine and coelenterazine n were obtained from Molecular Probes. SB202190, SB203580, SP600125, PD98059, U0126, KN-62, wortmannin, and anisomycin were from Tocris (Bristol, UK). SB220025, PD169316, ruthenium red, and the cellpermeable JNK inhibitory peptide were from Calbiochem (Bad Soden, Germany). Fugene was from Hoffmann-La Roche (Basel, Switzerland). Other reagents were from Sigma (Madrid, Spain) or Merck (Darmstadt, Germany). Structures of the p38 MAP kinase inhibitors used are as

follows: SB202190, 4-(4-fluorophenyl)-2-(4-hydroxyphenyl)-5-(4-pyridyl)-1H-imidazole; SB203580, 4-(4-fluorophenyl)-2-(4-methylsulfinylphenyl)-5-(4-pyridyl)-1H-imidazole, PD169316, 4-(4-fluorophenyl)-2-(4-nitrophenyl)-5-(4-pyridyl)-1H-imidazole; and SB220025, 4(4-fluorophenyl)-1-(4-piperidinyl)-5-(2-amino-4-pyrimidinyl)-imidazole. RESULTS Size of the agonist-induced [Ca2+]M peaks in HeLa cells In HeLa cells, activation of receptors for histamine or carbachol induces production of inositol 1,4,5-trisphosphate, release of Ca2+ from the endoplasmic reticulum (ER), and fast increase of [Ca2+] in mitochondria (15, 16). Most of the [Ca2+]M experiments that we carried out were performed with a HeLa cell clone (MM5) expressing mitmutAEQ, although similar results were obtained using transiently transfected wild-type HeLa cells. Mutated aequorin was used to reduce aequorin consumption and thus measure better the real changes in [Ca2+]M, avoiding the artifacts due to local saturation with Ca2+ of aequorin. To understand why we use this low-Ca2+affinity aequorin, it is necessary to know that previous measurements of histamine-induced [Ca2+]M peaks in HeLa cells have provided largely different results depending on the technique used. In the original work using high-affinity aequorin (15, 16), [Ca2+]M peaks of ∼3 µM were obtained. These peaks consumed nearly 30% of aequorin, and subsequent stimulation with either histamine or other agonists evoked much smaller responses (16). These results originated the concept of heterogeneous [Ca2+]M response at the subcellular level, resulting from the presence of “high-responsive” mitochondria (∼30% of the mitochondrial space) located close to the ER and “low-responsive” mitochondria placed away from the Ca2+-release sites. Once the aequorin present in the high-responsive mitochondria had been consumed, subsequent stimulation would give similar [Ca2+]M increases but much smaller light responses. Regarding the actual [Ca2+]M reached after histamine stimulation, these early data showed that 30% of the mitochondrial pool reaches [Ca2+]M values high enough to saturate high-affinity aequorin, but the real figure could not be measured. More recently, two groups have used different techniques to measure histamine-induced [Ca2+]M changes in HeLa cells. Collins et al. (17) used rhod-2 to measure [Ca2+]M in HeLa cells and performed confocal studies that showed [Ca2+]M peaks of ∼3.5 µM in response to histamine. Given that this value is six- to sevenfold greater than the Kd of rhod-2 for Ca2+ (thus close to saturation) and taking into account the heterogeneity in mitochondrial response, the possibility that a portion of the mitochondrial population may be suffering much larger increases in [Ca2+]M cannot be excluded. A different technique was used by Arnaudeau et al. (18). They targeted cameleons with different Ca2+ affinities to the mitochondria of HeLa cells and obtained quite different results depending on the affinity of the probe. When they used a high-affinity cameleon, the histamine-induced [Ca2+]M peak was ∼3 µM and nearly 25% of the pixels were saturated (a value close to the 30% of high-affinity aequorin consumed). With an intermediate-affinity cameleon, they obtained peaks of 49 µM. Finally, when the lowest affinity cameleon was used, the histamine-induced [Ca2+]M peak reached 106 µM and only 3% of the pixels were saturated. We have used here two different aequorin probes targeted to the mitochondria: mutated aequorin reconstituted with wild-type coelenterazine, which is able to report [Ca2+] up to 100 µM, and in

most of the experiments, mutated aequorin reconstituted with coelenterazine n. This probe keeps Ca2+ sensitivity up to the millimolar range and has been previously used to measure both steadystate [Ca2+] in the ER at ∼500 µM (14) and the [Ca2+]M transients reaching peak levels at ∼500µM, which were obtained in chromaffin cells after high-K+ or caffeine stimulation (7). Using any of these probes in MM5 cells, we obtained [Ca2+]M peaks induced by histamine or carbachol at ∼20–30 µM. Of course, this is the mean [Ca2+]M of the whole mitochondrial population, and surely there is a large subcellular variability in the [Ca2+]M responses of different mitochondria above and below this figure. SB202190 strongly enhances agonist-induced [Ca2+]M increase Figure 1 shows that the mitochondrial [Ca2+] peak induced by agonists in MM5 cells was dramatically increased when the cells were preincubated for a short time with SB202190, a specific inhibitor of p38 MAP kinase (19–21). In the experiments shown in the upper panel, mitmutAEQ was reconstituted with wild-type coelenterazine. Under control conditions, histamine and carbachol produced [Ca2+]M peaks of 24 ± 2 µM (n=6) and 28 ± 2 µM (n=6), respectively. In the presence of SB202190, the peaks increased to 95 ± 4 µM (n=6) and 90 ± 6 µM (n=6), respectively. This large increase in mitochondrial [Ca2+] strongly accelerated consumption of aequorin. Nearly 50% of aequorin was consumed during a single histamine or carbachol [Ca2+]M peak (45 ± 2%, n=6, and 46 ± 4% for histamine and carbachol, respectively). To evaluate better these large [Ca2+]M peaks, we reconstituted mitmutAEQ with coelenterazine n, which slows down the rate of consumption by about one order of magnitude (7, 14). The lower panels of Figure 1 show experiments similar to those of the upper panels but performed with this type of reconstituted aequorin. In this case, the control histamine and carbachol peaks reached 22 ± 1 µM (n=6) and 27 ± 1 µM (n=6), respectively, and they were increased in the presence of SB202190 to 75 ± 2 µM (n=6) and 76 ± 5 µM (n=6), respectively. These values were not affected by aequorin consumption, which was kept at low levels (the [Ca2+]M peak of histamine or carbachol in the presence of SB202190 consumed only 8.6 ± 0.2% or 10.3 ± 0.6% of aequorin, respectively). Therefore, aequorin consumption does not significantly affect the [Ca2+]M peak levels observed in the presence of SB202190. However, it is apparent from the figure that the [Ca2+]M peaks in the presence of SB202190 (but not in the absence) were wider when coelenterazine n was used to reconstitute mitmutAEQ, suggesting that the faster decrease of the [Ca2+]M seen in the experiments of the upper panels was in part due to aequorin consumption. In conclusion, although aequorin consumption does not modify the height of the peaks, reconstitution with coelenterazine n is required to obtain the real kinetics of the [Ca2+]M peaks in the presence of SB202190. From a kinetic point of view, the increase in the height of the [Ca2+]M peaks in the presence of SB202190 was due to an acceleration of the rate of Ca2+ entry into mitochondria. In experiments similar to those of Figure 1, lower panels, the rate of increase in [Ca2+]M induced by histamine jumped from 2.63 ± 0.2 µM/s (n=6) in the controls to 11.3 ± 0.6 µM/s (n=6) in the presence of SB202190, and that induced by carbachol increased from 1.7 ± 0.1 µM/s (n=6) in the controls to 6.5 ± 0.6 µM/s (n=6) in the presence of the inhibitor.

Some experiments were also conducted using a HeLa cell clone expressing wild-type aequorin reconstituted with wild-type coelenterazine (data not shown). It is remarkable that even though this kind of aequorin becomes saturated and rapidly consumed at [Ca2+] below 10 µM, the total aequorin consumption obtained after stimulation with histamine or carbachol in the presence of SB202190 was only 63 ± 2% (n=4). This suggests that nearly 40% of the mitochondria are excluded from the large increase in [Ca2+]M induced by SB202190. As reported previously, this heterogeneity probably depends on the precise subcellular location of mitochondria with respect to the Ca2+ release sites in the ER (15, 16). SB202190 reduces agonist-induced [Ca2+]c increase The increased mitochondrial Ca2+ uptake induced by SB202190 could be due to a direct activation of Ca2+ entry into mitochondria but could also be secondary to an increase in the [Ca2+]c peak induced by the agonists. This possibility was excluded by measuring [Ca2+]c in HeLa cells transfected with cytosolic aequorin. Figure 2 shows that the [Ca2+]c peaks induced by histamine or carbachol were not increased in the presence of SB202190. In fact, the [Ca2+]c peaks showed a significantly faster return of Ca2+ to resting levels in the presence of SB202190, suggesting that this compound somehow facilitates a faster extrusion of Ca2+ from the cytosol. This effect clearly suggests that the faster Ca2+ uptake by mitochondria is the primary effect of SB202190. Then, the activated mitochondrial Ca2+ uptake helps to damp faster the [Ca2+]c signal. Consistent with this hypothesis is the fact that SB202190 produced the same effects on the histamine-induced [Ca2+]M peak in Ca2+-free (EGTA-containing) medium (data not shown). Therefore, extracellular Ca2+ played no role in the effects of this compound. In addition, using a HeLa cell clone expressing aequorin targeted to the ER (13, 14), we could not detect any difference in the effects of histamine either in the presence or in the absence of SB202190 (data not shown). SB202190 does not affect mitochondrial Ca2+ release [Ca2+]M is determined by the opposite fluxes of Ca2+ in and out the mitochondria. Ca2+ enters through the Ca2+ uniporter and is extruded through exchangers, mainly a Na+/Ca2+ exchanger. Under resting conditions, the mitochondrial Ca2+ uniporter remains closed due to its low [Ca2+]c affinity, and the mitochondrial Na+/Ca2+ exchanger dominates, keeping [Ca2+]M very low. Instead, after cell stimulation, the uniporter is strongly activated by the increase in [Ca2+]c, and mitochondrial Ca2+ uptake becomes much faster than Ca2+ release. For this reason, inhibition of Ca2+ release has been shown to have little impact on the rate of [Ca2+]M increase or the height of the [Ca2+]M peaks after cell stimulation (7, 22). To completely exclude that the effects of SB202190 could be mediated by inhibition of mitochondrial Ca2+ release, we have used CGP37157. This compound is a specific inhibitor of the mitochondrial Na+/Ca2+ exchange, the main pathway for Ca2+ exit from mitochondria (9). Figure 3A shows that CGP37157 reduced the rate of Ca2+ release from mitochondria after histamine stimulation but did not increase the size of the [Ca2+]M peak. This clearly shows that the effects of SB202190 do not match those of an inhibition of Ca2+ release. In addition, Figure 3B shows that the effects of both SB202190 and CGP37157 were additive. In cells treated with SB202190, addition of CGP37157 also strongly reduced the rate of Ca2+ release from mitochondria after histamine stimulation.

The effects of SB202190 appear fast and are rapidly reversible The results previously described clearly indicate that the increase in the rate of mitochondrial Ca2+ uptake induced by SB202190 (Fig. 1) is due to a direct stimulation of mitochondrial Ca2+ entry. Given that this compound is considered a specific and selective inhibitor of p38 MAP kinase, our working hypothesis was that a constitutively active p38 MAP kinase would keep phosphorylated some component of the mitochondrial Ca2+ uptake machinery, leading to a tonic inhibition of this pathway. Inhibition of p38 with SB202190 would lead to dephosphorylation and activation of mitochondrial Ca2+ entry. Washout of this compound should then lead again to rephosphorylation and inhibition of Ca2+ entry. We decided to explore first the time course of these phenomena by studying the time required for SB202190 to produce its effect and the time required to relieve the inhibition. Figure 4 shows that both phenomena were quite fast. The increase in the [Ca2+]M peak induced by SB202190 was already evident in 10-fold, see Fig. 6). When the [Ca2+]c was further increased (Fig. 5C), the rate of mitochondrial Ca2+ uptake under control conditions increased rapidly, and the activation by SB202190 became smaller the higher the cytosolic [Ca2+]. In fact, at a [Ca2+]c of 5.5 µM, there was already no significant difference between the rates of mitochondrial Ca2+ uptake in the presence and in the absence of SB202190. The results of a series of experiments like those described in Figure 5 are resumed in Figure 6, which shows the dependence of the rate of mitochondrial Ca2+ uptake with [Ca2+]c both in the presence and in the

absence of SB202190. It can be clearly seen that SB202190 produced a large increase of that rate but only at relatively low [Ca2+]c (1–4 µM), precisely within the physiological range of [Ca2+]c. Mechanism of action of SB202190 To investigate the target on which this compound may be acting, we first studied the dependence of the effect with the concentration of SB202190. Figure 7 shows the effect of different concentrations of SB202190 on mitochondrial Ca2+ uptake in permeabilized cells (upper panel) and the dose-response curve (lower panel) calculated from a series of experiments like those of the upper panel. Concentrations as low as 1 µM already produced a clear stimulatory effect, although the maximal one was only obtained at 10–20 µM. In addition, Figure 7 also shows that SB202190-stimulated mitochondrial Ca2+ uptake was blocked in the presence of ruthenium red, suggesting that the stimulated Ca2+ flux was mediated by the Ca2+ uniporter. SB202190 is considered to be highly specific for the isoforms α and β of p38 MAP kinase at the concentrations used (21). However, inhibitors of p38 MAP kinase have been reported to inhibit also some isoforms of c-jun N-terminal kinase (JNK) at somewhat higher concentrations. To obtain further information regarding the protein kinase involved, we tested the effect of other inhibitors of the different MAP kinase pathways on the histamine-induced [Ca2+]M peaks. The results obtained are summarized in Table 1. The most surprising finding was that other inhibitors of p38 MAP kinase, such as PD169316, SB220025, or SB203580, which have a very close structural relationship to SB202190 (see Materials and Methods), had little or no effect even at higher concentrations. The involvement of JNK was tested by using both the specific inhibitor SP600125 and a cell-permeable JNK inhibitory peptide (23). This inhibitory peptide is linked to the 10-amino-acid HIV-TAT sequence that rapidly translocates inside many cell types, including HeLa (24). Incubation of HeLa cells with 10–20 µM of this peptide for 30 min did not modify the height of the histamine-induced [Ca2+]M peak. No effect was also obtained after treating the cells with 40 µM SP600125, suggesting that JNK is not involved. We have also tested the involvement of the classical MAP kinase pathway (ERK1/2) by using the inhibitors U0126 and PD98059. Again, preincubation of MM5 cells with either 50 µM U0126 or 50 µM PD98059 had little or no effect on the histamine-induced [Ca2+]M peaks. Of all the MAP kinase inhibitors tested, only SB203580 and U0126 produced some increase in the histamine-induced [Ca2+]M peaks (see Table 1), but it was very small compared with the fourfold increase induced by SB202190, and at much higher concentrations. Finally, no effect on the histamine-induced [Ca2+]M peaks was obtained after treating the cells for 5 min with 1 µM wortmannin, an inhibitor of the phophoinositide (phosphatidylinositol) 3-kinase, or with 10 µM KN62, an inhibitor of the Ca2+-calmodulin protein kinase II. Incubation of MM5 cells for 10–30 min with 10 µM anisomycin, a nonspecific activator of p38/JNK MAP kinases, did not modify the histamineinduced [Ca2+]M peaks. DISCUSSION We show a novel regulatory mechanism of mitochondrial Ca2+ uptake that may have physiological relevance for fine tuning of both mitochondrial and cytosolic [Ca2+] after cell stimulation. Our data show that the mitochondrial Ca2+ uniporter can increase its activity at

constant [Ca2+]c by more than one order of magnitude, rapidly and reversibly, just by adding or removing the p38 MAP kinase inhibitor SB202190. The effects of this compound can be explained only by an activation of the uniporter for several reasons: 1) Its effects were not due to any increase in Ca2+ entry from the extracellular medium or Ca2+ release from the ER (In fact, the histamine-induced [Ca2+]c transient was reduced in the presence of SB202190 [Fig. 2]). Therefore, the increased mitochondrial Ca2+ uptake induced by this compound was not secondary to any increase in [Ca2+]c.); 2) the effects of SB202190 cannot be explained by inhibition of mitochondrial Ca2+ release (Fig. 3); and 3) changes in membrane potential cannot explain the effects of SB202190. The Goldman-Hodgkin-Katz flux equation predicts that to increase the rate of mitochondrial Ca2+ uptake by 10-fold, the mitochondrial membrane potential should be increased by more than 10 times over the resting value. In addition, changes in membrane potential would affect Ca2+ uptake at every [Ca2+]c, whereas we find here that the stimulation took place only at low [Ca2+]c (1–4 µM). SB202190 thus increases the activity of the mitochondrial Ca2+ uniporter, and does it only at low [Ca2+]c (1–4 µM), precisely those more important from a physiological point of view. Thus, when this mechanism is activated, the increased mitochondrial Ca2+ uptake may damp [Ca2+]c changes, inhibiting [Ca2+]c-dependent processes. In fact, in intact HeLa cells, the increased activity of the Ca2+ uniporter induced by SB202190 led to a large increase in the mitochondrial Ca2+ peaks (Fig. 1) and to a concomitant decrease in the cytosolic [Ca2+] peaks (Fig. 2). Therefore, the increase in mitochondrial Ca2+ uptake was able to modulate [Ca2+]c. To fully appreciate this effect, it has to be taken into account that both the [Ca2+]M and the [Ca2+]c peaks have a similar time course, so that most of the Ca2+ accumulated in mitochondria is being released back into the cytosol before the end of the [Ca2+]c peak. Moreover, local [Ca2+]c changes in locations close to mitochondria are expected to be reduced in a much greater extent. SB202190 belongs to a family of pyridinyl imidazole compounds that have been shown to be very potent and specific inhibitors of p38α and p38β at the concentrations used (19–21). They bind competitively to the ATP pocket, and three amino acids in that pocket, particularly Thr106, determine the selectivity (19, 20). Changes at this position explain why p38α, p38β, and p38β2 are sensitive, but not p38γ and p38δ. In addition, this series of compounds has been widely used to evaluate the mechanism of physiological events that depend on the activity of p38 MAP kinase (19, 20). Therefore, the new regulation of the Ca2+ uniporter described here could be controlled by this pathway. However, the lack of effect of several structurally related p38 MAP kinase inhibitors (SB203580, PD169316, SB220025) suggests that a different but closely related protein kinase may be involved. However, the striking difference between the effects of SB202190 and these compounds, which are very similar from a structural point of view, indicates that SB202190 binds to its target with high specificity. The participation of some other related protein kinases was excluded by using inhibitors of other MAP kinases (PD98059, U0126, SP600125, JNK inhibitory peptide) and inhibitors of phophoinositide (phosphatidylinositol) 3-kinase (wortmannin) or Ca2+-calmodulin protein kinase II (KN-62). None of them produced significant effects on mitochondrial Ca2+ uptake. Although the protein kinase responsible of the effects of SB202190 on mitochondrial Ca2+ transport may not be p38 MAP kinase, it probably has a close relationship and may share some

of the properties of stress-activated kinases. Both p38 and JNK were originally described as stress-activated protein kinases because they were activated by osmotic or heat stress, but it has been shown more recently that they (and other MAP kinase pathways) can also be activated by other stimuli, including agonists acting through heterotrimeric G proteins (19, 20, 25). In addition, although many of the substrates of these kinases are in the nucleus, they have also been shown to phosphorylate cytosolic and even mitochondrial substrates. For example, mitochondrial bcl-2 is rapidly phosphorylated by JNK (26), and p38 has been involved in Bax translocation to mitochondria (27) and in the fast regulation of several plasma membrane ionic channel activities. In particular, recent evidence suggests that p38 may modulate the Na+/H+ exchanger (NHE-1) in vascular smooth cells (28), the N-type Ca2+ channel in a neuronal cell line (29), the acid secretion from gastric parietal cells (30), and the nonselective cation channel responsible for volume regulation in hepatocytes (31). Interestingly, in all the cases, p38 appeared to have an inhibitory role, as observed here. In addition, inhibition by p38 of the Na+-permeable channel of hepatocytes was constitutively active, as observed here for the mitochondrial Ca2+ uniporter, and this tonic inhibition appeared to play an essential role in the maintenance of cell volume (31). Constitutive activity of p38 MAP kinase also has been shown in breast cancer cells, where endogenous p38 MAP kinase activity correlates with invasiveness (32), and a remarkably high activity is also present in the normal adult brain (33, 34). In HeLa cells, p38 activity can be stimulated by agents such as anisomycin, but a considerable basal activity is also observed (35). Another important common property of the previously mentioned effects of p38 on transport systems is their fast timecourse. In the case of the volume-sensitive channels of hepatocytes, p38 inhibitors activated the channels without measurable delay (31). In the other cases, signaling by the agonists that stimulate p38 (bradykinin, angiotensin II, or carbachol) started immediately after the addition of the agonist (29) and produced maximum effect in a few minutes (28, 30). In conclusion, our results suggest that an SB202190-sensitive protein kinase keeps a tonic inhibition of the mitochondrial Ca2+ uniporter in HeLa cells, which is rapidly and reversibly relieved by this compound. In addition, the fact that the effect of SB202190 develops quickly in permeabilized cells, and is also rapidly reversible, provides important insights regarding the mechanism of the effect. Apart from excluding the participation of soluble cytosolic factors, these data suggest that all the elements responsible for the regulation are tightly attached to mitochondria. Further work will be necessary to identify the precise mechanism of this regulation and how it can be modulated under physiological conditions. Regarding the kinetic mechanism of the modulation of the Ca2+ uniporter, the observed shift in the [Ca2+]c activation curve suggests that the protein kinase activity may be reducing the [Ca2+]c affinity of the uniporter. It has been recently reported that mitochondrial Ca2+ uptake desensitizes for a period of 10–20 min after a pulse of histamine in HeLa cells (17). The mechanism described here could be responsible for that effect if histamine could stimulate still further the relevant protein kinase activity. However, in our hands, stimulation with histamine produced no effect on the rate of mitochondrial Ca2+ uptake measured in MM5 cells permeabilized immediately after stimulation (data not shown). Our finding opens a large new field of interactions among plasma membrane receptors and mitochondrial Ca2+ uptake. There is considerable evidence in different cell types that

mitochondrial Ca2+ uptake modulates [Ca2+]c transients (1–3). Therefore, mitochondrial Ca2+ uptake may modulate a large number of [Ca2+]c-dependent processes such as secretion, neurotransmission, and cell contraction. In chromaffin cells, for example, we and others have shown previously that catecholamine secretion is strongly enhanced when mitochondrial Ca2+ uptake is abolished (7, 36). The question was then how this mitochondrial Ca2+ uptake could be modulated to have physiological significance. The new regulatory pathway described here may be the link required to allow mitochondria performing a flexible regulation of [Ca2+]c-dependent phenomena. We can now predict, for example, that stimulation of this protein kinase pathway in neuronal cells should lead to inhibition of mitochondrial [Ca2+] uptake, local increase in [Ca2+]c, and potentiation of secretion. Agonists acting through this pathway may therefore prime the secretory response by blocking mitochondrial Ca2+ buffering. In addition, we should mention that several inhibitors of p38 MAP kinase now are being used in a series of preclinical and clinical studies as possible oral therapeutic agents to treat several inflammatory and cardiovascular diseases (19, 20, 37, 38). The new effect we show here of SB202190 on mitochondrial Ca2+ uptake may be highly relevant to interpret some of the clinical effects of these drugs. ACKNOWLEDGMENTS This work was supported by grants from the Spanish Ministry of Science and Technology (PM98/0142 and BFI2002-01397) and from Junta de Castilla y León (VA 005/02). REFERENCES 1. Duchen, M.R. (2000) Mitochondria and calcium: from cell signalling to cell death. J. Physiol. 529.1, 57–68 2. Rizzuto, R., Bernardi, P., and Pozzan, T. (2000) Mitochondria as all-round players of the calcium game. J. Physiol. 529.1, 37–47 3. Pozzan, T., and Rizzuto, R. (2000) The renaissance of mitochondrial calcium transport. Eur. J. Biochem. 267, 5269–5273 4. Park, Y.B., Herrington, J., Babcock, D.F., and Hille, B. (1996) Ca2+ clearance mechanisms in isolated rat adrenal chromaffin cells. J. Physiol. 492.2, 329–346 5. White, R.J., and Reynolds, I.J. (1997) Mitochondria accumulate Ca2+ following intense glutamate stimulation of cultured rat forebrain neurones. J. Physiol. 498.1, 31–47 6. Villalobos, C., Nuñez, L., Montero, M., García, A. G., Alonso, M.T., Chamero, P., Alvarez, J., and García-Sancho, J. (2002) Redistribution of Ca2+ among cytosol and organella during stimulation of bovine chromaffin cells. FASEB J. 16, 343–353 7. Montero, M., Alonso, M.T., Carnicero, E., Cuchillo-Ibañez, I., Albillos, A., Garcia, A.G., Garcia-Sancho, J., and Alvarez, J. (2000) Chromaffin-cell stimulation triggers fast millimolar mitochondrial Ca2+ transients that modulate secretion. Nature Cell. Biol. 2, 57–61

8. Gunter, T.E., Buntinas, L., Sparagna, G.C., and Gunter, K.K. (1998) The Ca2+ transport mechanisms of mitochondria and Ca2+ uptake from physiological-type Ca2+ transients. Biochim. Biophys. Acta 1366, 5–15 9. Bernardi, P. (1999) Mitochondrial transport of cations: channels, exchangers and permeability transition. Physiol. Rev. 79, 1127–1155 10. Rustenbeck, I., Eggers, G., Reiter, H., Munster, W., and Lenzen, S. (1998) Polyamine modulation of mitochondrial calcium transport. Biochem. Pharmacol. 56, 977–985 11. Allen, D.G., Blinks, J.R., and Prendergast, F.G. (1971) Aequorin luminescence: relation of light emission to calcium concentration--a calcium-independent component. Science 195, 996– 998 12. Brini, M., Marsault, R., Bastianutto, C., Alvarez, J., Pozzan, T., and Rizzuto, R. (1995) Transfected aequorin in the measurement of cytosolic Ca2+ concentration ([Ca2+]c). A critical evaluation. J. Biol. Chem. 270, 9896–9903 13. Montero, M., Brini, M., Marsault, R., Alvarez, J., Sitia, R., Pozzan, T., and Rizzuto, R. (1995) Monitoring dynamic changes in free Ca2+ concentration in the endoplasmic reticulum of intact cells. EMBO J. 14, 5467–5475 14. Montero, M., Barrero, M.J., and Alvarez, J. (1997) [Ca2+] microdomains control agonistinduced Ca2+ release in intact HeLa cells. FASEB J. 11, 881–885 15. Rizzuto, R., Brini, M., Murgia, M., and Pozzan, T. (1993) Microdomains with high Ca2+ close to IP3-sensitive channels that are senses by neighboring mitochondria. Science 262, 744– 747 16. Rizzuto, R., Bastianutto, C., Brini, M., Murgia, M., and Pozzan, T. (1994) Mitochondrial Ca2+ homeostasis in intact cells. J. Cell Biol. 126, 1183–1194 17. Collins, T.J., Lipp, P., Berridge, M.J., and Bootman, M.D. (2001) Mitochondrial Ca2+ uptake depends on the spatial and temporal profile of cytosolic Ca2+ signals. J. Biol. Chem. 276, 26411– 26420 18. Arnaudeau, S., Kelley, W.L., Walsh, J.V. Jr., and Demaureux, N. (2001) Mitochondria recycle Ca2+ to the endoplasmic reticulum and prevent the depletion of neighboring endoplasmic reticulum regions. J. Biol. Chem. 276, 29430–29439 19. Herlaar, E., and Brown, Z. (1999) P38 MAPK signalling cascades in inflammatory disease. Mol. Med. Today 5, 439–447 20. Lee, J.C., Kumar, S., Griswold, D.E., Underwood, D.C., Votta, B.J., and Adams, J.L. (2000) Inhibition of p38 MAP kinase as a therapeutic strategy. Immunopharmacol. 47, 185–201

21. Davies, S.P., Reddy, H., Caivano, M., and Cohen, P. (2000) Specificity and mechanism of action of some commonly used protein kinase inhibitors. Biochem. J. 351, 95–105 22. Babcock, D.F., Herrington, J., Park, Y.-B., and Hille, B. (1997) Mitochondrial participation in the intracellular Ca2+ network. J. Cell. Biol. 136, 833–843 23. Bonny, C. Oberson, A., Negri, S., Sauser, C., and Schorderet, D.F. (2001) Cell-permeable peptide inhibitors of JNK. Diabetes 50, 77–82 24. Vives, E., Brodin, P., and Lebleu, B. (1997) A truncated HIV-1 Tat protein basic domain rapidly translocates through the plasma membrane and accumulates in the cell nucleus. J. Biol. Chem. 272, 16010–16017 25. Ono, K., and Han, J. (2000) The p38 signal transduction pathway. Activation and function. Cellular Signalling 12, 1–13 26. Kharbanda, S., Saxena, S., Yoshida, K., Pandey, P., Kaneki, M., Wang, Q., Cheng, K., Chen, Y.N., Campbell, A., Sudha, T., Yuan, Z.M., Narula, J., Weichselbaum, R., Nalin, C., and Kufe, D. (2000) Translocation of SAPK/JNK to mitochondria and interaction with Bcl-xL in response to DNA damage. J. Biol. Chem. 275, 322–327 27. Ghatan, S., Larner, S., Kinoshita, Y., Hetman, M., Patel, L., Xia, Z., Youle, R.J., and Morrison, R.S. (2000) p38 MAP kinase mediates Bax translocation in nitric oxide–induced apoptosis in neurons. J. Cell. Biol. 150, 335–347 28. Kusuhara, M., Takahashi, E., Peterson, T.E., Abe, J., Ishida, M., Han, J., Ulevitch, R., and Berk, B.C. (1998) P38 kinase is a negative regulator of angiotensin II signal transduction in vascular smooth muscle cells. Effects on Na+/H+ exchange and ERK1/2. Circ. Res. 83, 824–831 29. Wilk-Blaszczak, M.A., Stein, B., Xu, S., Barbosa, M.S., Cobb, M.H., and Belardetti, F. (1998) The mitogen-activated protein kinase p38-2 is necessary for the inhibition of the N-type current by bradykinin. J. Neurosci. 18, 112–118 30. Pausawasdi, N., Ramamoorthy, S., Stepan, V., del Valle, J., and Todisco, A. (2000) Regulation and function of p38 protein kinase in isolated canine gastric parietal cells. Am. J. Physiol. Gastrointest. Liver Physiol. 278, G24–G31 31. Feranchak, A.P., Berl, T., Capasso, J., Wojtaszek, P.A., Han, J., and Fitz. J.G. (2001) P38 MAP kinase modulates liver cell volume through inhibitin of membrane Na+ permeability. J. Clin. Invest. 108, 1495–1504 32. Huang, S., New, L., Pan, Z., Han, J., and Nemerow, G.R. (2000) Urokinase plasminogen activator/urokinase-specific surface receptor expression and matrix invasion by breast cancer cells requires constitutive p38α mitogen-activated protein kinase activity. J. Biol. Chem. 275, 12266–12272

33. Mielke, K., and Herdegen, T. (2000) JNK and p38 stresskinases – degenerative effectors of signal-transduction-cascades in the nervous system. Prog. Neurobiol. 61, 45–60. 34. Lee, S.H., Park, J., Che, Y., Han, P.L., and Lee, J.K. (2000) Constitutive activity and differential localization of p38α and p38β MAPKs in adult mouse brain. J. Neurosci. Res. 60, 623–631 35. Sayed, M., Kim., S.O., Sahl, B.S., Issinger, O.-G., and Pelech, S.L. (2000) Stress-induced activation of protein kinase CK2 by direct interaction with p38 mitogen-activated protein kinase. J. Biol. Chem. 275, 16569–16573 36. Giovannucci, D.R., Hlubek, M.D., and Stuenkel, E.L. (1999) Mitochondria regulate the Ca2+exocytosis relationship of bovine adrenal chromaffin cells. J. Neurosci. 19, 9261–9270 37. Behr, T.M., Nerurkar, S.S., Nelson, A.H., Coatney, R.W., Woods, T.N., Sulpizio, A., Chandra, S., Brooks, D.P., Kumar, S., Lee, J.C., Ohlstein, E.H., Angermann, C.E., Adams, J.L., Sisko, J., Sackner-Bernstein, J.D., and Willette, R.N. (2001) Hypertensive end-organ damage and premature mortality are p38 mitogen-activated protein kinase-dependent in a rat model of cardiac hypertrophy and dysfunction. Circulation 104, 1292–1298 38. Sugden, P.H., and Clerk, A. (1998) “Stress-responsive” mitogen-activated protein kinases (cjun N-terminal kinases and p38 mitogen-activated protein kinases) in the myocardium. Circ. Res. 83, 345–352 Received June 21, 2002; accepted August 21, 2002.

Table 1 Effects of inhibitors of different mitogen-activated protein kinase pathways on the histamine-induced [Ca2+]M increasea Inhibitor PD169316, 30 µM SB220025, 30 µM SB203580, 30 µM SP600125, 40 µM JNK inhibitor peptide,10–20 µM PD98059, 50 µM U0126, 50 µM

[Ca2+]M increase (% of control) 92±7% (n=6) 82±7% (n=4) 157±13% (n=6) 99±8% (n=6) 92±11% (n=12) 94±10% (n=6) 144±14% (n=6)

Experiments similar to those shown in Figure 1 were performed in the presence of the different inhibitors. Cells were preincubated with the inhibitors for 5 min, except in the case of the JNK inhibitory peptide, which was added 30 min before stimulation. Percentages indicate the height of the histamine-induced [Ca2+]M peaks in the presence of the inhibitors with respect to that in the controls. Data are mean ±SE a

Fig. 1

Figure 1. Effect of SB202190 on the [Ca2+]M peaks induced by histamine and carbachol. In the upper panel, MM5

cells expressing mitmutAEQ reconstituted with wild-type coelenterazine were stimulated either with 100 µM histamine or 500 µM carbachol, as indicated. In the experiments labeled “+SB,” 10 µM SB202190 was present during stimulation and also for 5 min before the stimulus. In the lower panel, similar experiments were performed using MM5 cells reconstituted with coelenterazine n. Experiments shown are representative of six similar ones of each type.

Fig. 2

Figure 2. Effect of SB202190 on the [Ca2+]c peaks induced by histamine and carbachol. HeLa cells expressing

cytosolic aequorin reconstituted with wild-type coelenterazine were stimulated with 100 µM histamine or 500 µM carbachol, as indicated. In the experiments labeled “+SB,” 10 µM SB202190 was present during stimulation and also for 5 min before the stimulus. The traces shown are the average of four similar experiments of each type.

Fig. 3

Figure 3. Effect of CGP37157 on the histamine-induced [Ca2+]M peaks both in the absence and in the presence of

SB202190. Left panel: MM5 cells expressing mitmutAEQ reconstituted with coelenterazine n were stimulated with 100 µM histamine either in control cells or after preincubation for 2 min with 10 µM CGP37157 (curve marked “+CGP”). CGP37157 was also present during and after histamine stimulation. Right panel: Same as the left panel, but in both curves, 10 µM SB202190 was present 5 min before, during, and after histamine stimulation. Experiments shown are representative of six similar ones of each type.

Fig. 4

Figure 4. Time course of the effect and washout of SB202190. In the upper panel, MM5 cells expressing mitmutAEQ reconstituted with wild-type coelenterazine were stimulated with 100 µM histamine either in control cells or after preincubation with 10 µM SB202190 for different times, as indicated. Bars show the height of the histamine-induced [Ca2+]M peaks normalized using the height of the control peak as 100%. In the lower panel, MM5 cells expressing mitmutAEQ reconstituted with wild-type coelenterazine were incubated for 5 min with 10 µM SB202190 and then stimulated with 100 µM histamine after a variable period without the inhibitor, as indicated. Bars are also normalized using the height of the control peak as 100%. Error bars show the SE of four to five different experiments of each type.

Fig. 5

Figure 5. Effect of SB202190 on mitochondrial Ca2+ uptake in permeabilized cells. MM5 cells expressing mitmutAEQ reconstituted with coelenterazine n were permeabilized as described in Materials and Methods. Then buffers containing different [Ca2+] and/or 10 µM SB202190 were perfused, as indicated. In the lower panels, the traces obtained in the presence and in the absence of 10 µM SB202190 are superimposed for each [Ca2+]. In the traces labeled “+SB,” the inhibitor was present for 2 min before and during perfusion of the [Ca2+] buffer.

Fig. 6

Figure 6. Dependence of the rate of mitochondrial Ca2+ uptake on [Ca2+]c in the presence and in the absence of

SB202190. Mean uptake rate data obtained from experiments similar to those of Figure 4 are shown. Error bars show the of 4–10 different experiments of each condition.

SE

Fig. 7

Figure 7. Dose dependency of the increase in Ca2+ uptake rate induced by SB202190. In the upper panel, MM5 cells

reconstituted with coelenterazine n were permeabilized, and then 2µM [Ca2+] was perfused for 3 min. Then, different concentrations of SB202190, as indicated, were added to the same [Ca2+] buffer. In the trace labeled “RR+SB,” 10 µM SB202190 and 1 µM ruthenium red were both present 2 min before and during Ca2+ perfusion. The lower panel shows the dependency of the stimulation of mitochondrial Ca2+ uptake with the concentration of SB202190. Uptake rates are normalized using the control rate as 100% (horizontal line). Error bars show the SE of four different experiments of each condition.