Ca2+ Efflux in Mitochondria from the YeastEndomyces magnusii *

Calcium release pathways in Ca2+-preloaded mitochondria from the yeastEndomyces magnusii were studied. In the presence of phosphate as a permeant anion, Ca2+ was released from respiring mitochondria only after massive cation loading at the onset of anaerobiosis. Ca2+ release was not affected by cyclosporin A, an inhibitor of the mitochondrial permeability transition. Aeration of the mitochondrial suspension inhibited the efflux of Ca2+ and induced its re-uptake. With acetate as the permeant anion, a spontaneous net Ca2+ efflux set in after uptake of ∼150 nmol of Ca2+/mg of protein. The rate of this efflux was proportional to the Ca2+ load and insensitive to aeration, protonophorous uncouplers, and Na+ions. Ca2+ efflux was inhibited by La3+, Mn2+, Mg2+, tetraphenylphosphonium, inorganic phosphate, and nigericin and stimulated by hypotonicity, spermine, and valinomycin in the presence of 4 mm KCl. Atractyloside andt-butyl hydroperoxide were without effect. Ca2+efflux was associated with contraction, but not with mitochondrial swelling. We conclude that the permeability transition pore is not involved in Ca2+ efflux in preloaded E. magnusii mitochondria. The efflux occurs via an Na+-independent pathway, in many ways similar to the one in mammalian mitochondria.

Calcium plays an important role as an intracellular messenger in signal transduction (1,2). Mitochondrial Ca 2ϩ uptake is important in relaying a signal for stimulation of respiration and oxidative phosphorylation (3,4), which has encouraged study of mitochondrial Ca 2ϩ handling (5). Mitochondria are receiving increasing attention also due to their central role in the mechanism of cell death (6). Animal mitochondria take up Ca 2ϩ by a uniport mechanism and release it by a number of different mechanisms, including Na ϩ antiport, a sodium-independent pathway, and via opening of a cyclosporin A-sensitive pore (for review, see Ref. 7).
Yeast mitochondria generally lack a Ca 2ϩ uptake pathway of any significant physiological relevance (8,9). Instead, a vacuolar V-ATPase and a Ca 2ϩ /H ϩ antiporter regulate cytosolic [Ca 2ϩ ] (10). However, we have found that tightly coupled mitochondria from the yeast Endomyces magnusii are able to take up Ca 2ϩ by a uniport mechanism, although the apparent K m is rather high, 150 -180 M (11, 12). Ca 2ϩ uptake at low concentrations can, however, as in animal mitochondria, be substantially increased by micromolar concentrations of polyamines (13)(14)(15)(16)(17). Also ADP, Ca 2ϩ itself, and a high intramitochondrial NADH/NAD ϩ ratio stimulate mitochondrial Ca 2ϩ uptake (15)(16)(17)(18). In the presence of all these physiological modulators, the initial rate of Ca 2ϩ uptake is quite high (up to 2 mol/min/mg of protein), and the mitochondrial Ca 2ϩ -buffering capacity is remarkably high.
Ca 2ϩ uptake by E. magnusii mitochondria is thus potentially important in Ca 2ϩ homeostasis and signal transduction (15)(16)(17)(18). The Ca 2ϩ efflux pathways are also of interest in this context. In this study, we have characterized the mitochondrial Ca 2ϩ efflux mechanism, which shares many of its characteristics with the sodium-independent Ca 2ϩ efflux mechanism of mitochondria from non-excitable mammalian cells (19).

EXPERIMENTAL PROCEDURES
The yeast E. magnusii strain VKM Y261 was grown in glycerolcontaining semisynthetic medium as described previously (20). Cells were harvested at the late exponential growth phase (10 -13 g (wet weight)/liter). Mitochondria were isolated by the method developed in our laboratory (17). The oxygen consumption in mitochondrial suspensions was monitored polarographically with a Clark-type electrode in medium containing 0.6 M mannitol, 1 mM Tris phosphate (pH 7.4), 1 mM EDTA, 20 mM pyruvate, 5 mM malate, and mitochondria corresponding to 0.5 mg/ml protein. The mitochondrial preparations were well coupled, showing respiratory control ratios of ϳ4 under these conditions. ADP/oxygen ratios were close to the theoretical maximum. Respiratory control and ADP/oxygen ratios were calculated as described by Chance and Williams (21). Mitochondria were fully active for at least 4 h after preparation.
Ca 2ϩ uptake was assayed by the murexide method employing dual wavelength photometry at 507-540 nm with a Hitachi 557 spectrophotometer. Unless otherwise specified in the figure legends, the incubation medium contained 0.6 M mannitol, 2 mM Tris phosphate or 20 mM Tris acetate (pH 7.4), 20 mM Tris pyruvate, 5 mM malate, 50 M murexide, and mitochondria corresponding to 0.5 mg/ml protein. The inner mitochondrial membrane transmembrane potential (⌬⌿) 1 was measured at the wavelength pair 523/555 nm with 7 M safranin (22). Mitochondrial swelling was monitored by recording changes in absorbance at 540 nm. Protein was assayed by the method of Bradford (23) with bovine serum albumin as the standard.

RESULTS
Massive Ca 2ϩ Loading in the Presence of Phosphate as the Anion-In the presence of P i , yeast mitochondria were able to take up almost all of the Ca 2ϩ added in successive aliquots (up to 600 nmol of Ca 2ϩ /mg of protein) and to retain it until the oxygen was exhausted after 5-10 min of incubation (Fig. 1, trace a). Anaerobiosis induced a rapid collapse of ⌬⌿ (Fig. 1, trace c), the driving force of Ca 2ϩ uptake by the calcium uniporter. The rate of Ca 2ϩ uptake was much faster than that of efflux induced by anaerobiosis (Fig. 1), which is in accordance with the ability of the mitochondria to reduce the [Ca 2ϩ ] to very low levels at steady state. On the other hand, upon collapse of ⌬⌿, efflux could be by reversal of the uptake mechanism, the calcium uniporter. Intensive aeration of the incubation medium fully reestablished ⌬⌿, prevented spontaneous Ca 2ϩ release, and elicited re-uptake of the Ca 2ϩ released (Fig.  1, trace a). Aeration before each Ca 2ϩ addition also prevented Ca 2ϩ release ( Fig. 1, trace b) and a drop in ⌬⌿ (trace d), thus increasing the Ca 2ϩ -buffering capacity of the yeast mitochondria. The efflux rate was not affected by the addition of CsA (Fig. 1, trace a).
Spontaneous Efflux of Accumulated Ca 2ϩ in Acetate-containing Medium-Since no Ca 2ϩ efflux was observed in respiring yeast mitochondria in the presence of P i , acetate was used instead as a permeant anion. The Ca 2ϩ efflux pathways in animal mitochondria are not influenced by acetate and thus can be studied in the presence of this anion, the calcium salt of which is soluble (24). In yeast mitochondria, Ca 2ϩ efflux ensued after uptake of ϳ75 M Ca 2ϩ (Fig. 2, trace a). Ca 2ϩ release was not due to a decline in ⌬⌿ (Fig. 2, trace b), and it was insensitive to micromolar concentrations of CsA (data not shown). The efflux was thus spontaneous and not due to anaerobiosis or induction of the mitochondrial permeability transition (MPT). The rate of Ca 2ϩ efflux was proportional to the Ca 2ϩ load (Fig.  3A). It was inhibited half-maximally by P i at 0.3 mM (Fig. 3B).
Characterization of the Ca 2ϩ Efflux Pathway-In acetate medium, Ca 2ϩ uptake would cause mitochondrial swelling due to accumulation of calcium acetate in the matrix, whereas efflux would result in contraction. Fig. 4 shows the effect of various agents on the contraction due to spontaneous Ca 2ϩ efflux in mitochondria respiring on pyruvate ϩ malate. In mitochondria to which 100 M Ca 2ϩ had been added, the addition of N-ethylmaleimide only slightly slowed contraction, whereas t-butyl hydroperoxide, oxalacetate, atractyloside, and the uncoupling agent CCCP in the presence of EGTA were without any effect (Fig. 4).
In animal mitochondria, efflux of Ca 2ϩ can be studied by inhibiting the uniporter with ruthenium red (25,26). However, the E. magnusii mitochondrial calcium uniporter is not inhibited by ruthenium red (12) and may even be stimulated by it (16). In the acetate medium, there was a spontaneous net efflux of accumulated Ca 2ϩ (Fig. 2), but this did not exclude a simultaneous uptake. We therefore examined whether La 3ϩ , a competitive inhibitor of mitochondrial Ca 2ϩ transport (27) used as an inhibitor of the uniporter. However, although Ca 2ϩ uptake was sensitive to low concentrations of La 3ϩ (Fig. 5A,  trace b), efflux was also slightly inhibited, more so at higher concentrations (trace a). Half-maximal inhibition of Ca 2ϩ uptake was attained at ϳ25 M La 3ϩ . The effects of La 3ϩ were thus different for uptake and efflux of Ca 2ϩ .
Ca 2ϩ efflux was inhibited by Mn 2ϩ (Fig. 5B, trace a) and by Mg 2ϩ (Fig. 5C, trace a). These cations also affected the rate of Ca 2ϩ uptake and efflux with a different concentration dependence (Fig. 5, B and C). Thus, 200 M Mn 2ϩ inhibited Ca 2ϩ efflux by 80% (Fig. 5B, trace a), whereas Ca 2ϩ uptake was inhibited by only 50% (trace b). Mg 2ϩ at 1.5 mM almost totally blocked Ca 2ϩ efflux (Fig. 5C, trace a) and inhibited Ca 2ϩ uptake by only 50% (trace b). In the case of Mg 2ϩ (Fig. 5C), the concentration dependence of inhibition of Ca 2ϩ uptake and efflux was even more strikingly different. TPP ϩ is a potent and apparently specific inhibitor of the sodium-independent Ca 2ϩ efflux pathway in animal mitochondria (28). We also found that in yeast mitochondria, TPP ϩ strongly inhibited Ca 2ϩ efflux (Fig. 5D), with a maximal inhibitory effect at 10 M and a half-maximal effect at 1.8 M (calculated in Dixon plots), which is close to the K i value obtained for liver mitochondria (28). Uptake of Ca 2ϩ was inhibited only slightly (10%) by 10 M TPP ϩ (data not shown). Ca 2ϩ efflux in yeast mitochondria was almost totally inhibited by 1.2 mM P i (Fig. 3B), a concentration that was found to be optimal in sustaining Ca 2ϩ uptake (11).
Spermine increased the rate of Ca 2ϩ efflux when added both before (Fig. 6, trace c) and after (trace a) Ca 2ϩ , with no effect upon ⌬⌿ (traces a and b). The spermine-induced Ca 2ϩ release was nonlinear and similar to that seen after the addition of the Ca 2ϩ ionophore A23187 (Fig. 2). The addition of A23187 after spermine induced further Ca 2ϩ release (Fig. 6, trace a).
The rate of Ca 2ϩ efflux in yeast mitochondria was not affected by the redox state of pyridine nucleotides and was only slightly inhibited by micromolar concentrations of ADP (data not shown). Preincubation with 4 mM NADH to greatly reduce intramitochondrial pyridine nucleotides had no effect (data not shown), whereas these substances have been shown to stimulate Ca 2ϩ uptake (17).
Swelling of mammalian mitochondria in hypotonic medium was found to stimulate the sodium-independent Ca 2ϩ efflux pathway, presumably because of stretching of the inner mitochondrial membrane (24,34,35). Likewise, in yeast mitochondria, a significantly (2.5-fold) enhanced Ca 2ϩ efflux was observed when mitochondria were incubated in hypotonic medium (Fig. 7A, trace b).
To elucidate the mechanism of the efflux pathway, we examined the effects of CCCP (a protonophorous uncoupler), nigericin (an ionophore acting via K ϩ /H ϩ antiport), and valinomycin (a K ϩ ionophore). Ca 2ϩ efflux was only partially inhibited (30%) by 20 -200 nM CCCP (data not shown), whereas little effect (if any) was seen on the associated contraction of mitochondria (Fig. 4). Ca 2ϩ efflux was completely inhibited by 100 nM nigericin (the medium was supplemented with 0.5 mM KCl) (Fig. 7B, trace a). Nigericin, which equilibrates K ϩ and H ϩ gradients across the inner mitochondrial membrane and is thereby able to partially transform ⌬pH into ⌬⌿, slightly increased ⌬⌿ at most. The addition of 100 nM valinomycin (the medium was supplemented with 4 mM KCl) caused dissipation of ⌬⌿ and increased the rate of Ca 2ϩ efflux (Fig. 7B, trace b). DISCUSSION In the presence of P i , E. magnusii mitochondria are capable of massive uptake of Ca 2ϩ (see Fig. 1). In animal mitochondria, P i has been reported to induce the release of accumulated P i (36), which is now known to be due to induction of MPT (7). Massive accumulation was observed in animal mitochondria in the presence of both P i and ATP due to precipitation of calcium phosphates (for review, see Ref. 37) and inhibition of MPT (7). In the absence of adenine nucleotides, Ca 2ϩ uptake by rat liver mitochondria is also stimulated by P i , and the sodium-indepen-

FIG. 3. Effect of Ca 2؉ load (A) and of P i (B) on the rate of Ca 2؉ efflux in yeast mitochondria.
The medium was the same as that described in the legend to Fig. 2, but without murexide. In A and B, 100 and 500 M Ca 2ϩ were added, respectively. dent efflux is inhibited, which was interpreted as being due to lowering of matrix [Ca 2ϩ ] by its binding to P i (19). In P idepleted liver mitochondria, the sodium-independent Ca 2ϩ ef-flux rate is stimulated 15-fold over that in the presence of P i and is potently inhibited by the addition of P i (38). Stimulation of Ca 2ϩ uptake by P i and inhibition of efflux have also been described in microsomes and were also interpreted as being due to binding of Ca 2ϩ by P i (39). The same mechanism may account for the effect of P i on Ca 2ϩ efflux in E. magnusii mitochondria.
The efflux of accumulated Ca 2ϩ in acetate medium could occur by a reversal of the calcium uniporter upon lowering of ⌬⌿, by specific efflux mechanisms such as the Ca 2ϩ /2Na ϩ antiporter (40) and the Ca 2ϩ /nH ϩ antiporter (7,19,28) in animal mitochondria, or by reversible opening of the MPT pore. However, the efflux rate was not influenced by uncoupling amounts of CCCP (Fig. 4) or Na ϩ (data not shown), which excludes the uniporter and the Ca 2ϩ /2Na ϩ antiporter mechanisms. This is borne out by data showing that the uptake and efflux pathways differ in their sensitivities to La 3ϩ (Fig. 5A) and modulators such as Mn 2ϩ (Fig. 5B), Mg 2ϩ (Fig. 5C), TPP ϩ (Fig. 5D), and spermine (Fig. 6). Efflux was only slightly inhibited by ADP (data not shown), whereas uptake was potently stimulated (with a half-maximal effect at 3-5 M and at a maximal effect at 25 M) (15,17,18). In addition, the efflux rate was not influenced by preincubation of the mitochondrial suspension with 4 mM NADH, which ensures a high redox state of the intramitochondrial pyridine nucleotides in yeast mitochondria (41), although Ca 2ϩ uptake is stimulated under these conditions (17).
In animal mitochondria, ADP inhibits and atractyloside stimulates Ca 2ϩ release mediated by the MPT pore, presumably by affecting the conformation of the adenine nucleotide carrier (7). In E. magnusii mitochondria, Ca 2ϩ efflux was only slightly affected by ADP (data not shown). The addition of atractyloside, pro-oxidants, N-ethylmaleimide, or small amounts of CCCP or oxidation of pyridine nucleotides by oxalacetate had little effect (if any) on the rate of mitochondrial contraction associated with Ca 2ϩ efflux (Fig. 4). All these findings indicate that there is no Ca 2ϩ -stimulated opening of a MPT pore in these mitochondria. This conclusion is strongly supported by the absence of inhibition by CsA and by the massive accumulation of Ca 2ϩ in the presence of P i (Fig. 1) that would induce MPT in animal mitochondria (7). No MPT of this type has been found in Saccharomyces cerevisiae mitochondria (42).
Mg 2ϩ ions are believed to stabilize the inner mitochondrial membrane by modulating ion transport and inhibiting the Ca 2ϩ -activated phospholipase A 2 and MPT (43), the latter not being of relevance in the case of yeast mitochondria as dis-cussed above. The mechanism of the inhibitory action of Mn 2ϩ on Ca 2ϩ efflux has not been conclusively established. It seems likely that in E. magnusii mitochondria, Mg 2ϩ and Mn 2ϩ exert their action from the cytosolic side since these mitochondria have only a low-capacity transport system for Mg 2ϩ (44), and the calcium uniporter supports slow transport of Sr 2ϩ , Mn 2ϩ , or Ba 2ϩ only at high millimolar concentrations (data not shown). Furthermore, these cations exert their inhibitory action without any lag.
Spermine is well known as an important modulator of the Ca 2ϩ transport system in animal mitochondria, stimulating electrogenic Ca 2ϩ uptake (45)(46)(47)(48)(49) and mostly inhibiting Ca 2ϩ efflux (29,46,49). We have also reported a stimulatory effect of low physiological concentrations of spermine on Ca 2ϩ uptake by E. magnusii mitochondria (13)(14)(15)(16)(17). The mechanism of the stimulation of the efflux rate by spermine may be an allosteric change in the conformation of the transporter or/and an effect of binding of spermine to negatively charged membrane groups, thus favoring the binding and dissociation of Ca 2ϩ from the membrane. It is also possible that Ca 2ϩ itself stimulates the transporter by the same allosteric effect on the transporter, which would explain the stimulation of efflux after uptake over a certain threshold.
Nigericin inhibited Ca 2ϩ efflux in the presence of 0.5 mM K ϩ under conditions that did not change ⌬⌿, whereas valinomycin stimulated efflux in the presence of 4 mM K ϩ (Fig. 7B). Under these conditions, valinomycin caused a decrease in ⌬⌿ and swelling (Fig. 4). Since there should be no appreciable pH gradient in the presence of acetate, and a drop in ⌬⌿ induced by CCCP did not change the efflux rate (Fig. 4), it seems likely that the inhibition is due to reduction of the mitochondrial volume. The valinomycin-induced stimulation of efflux could then be due to swelling (Fig. 4) induced by accumulation of potassium acetate in the matrix.
Taken together, these data indicate that Ca 2ϩ efflux in E. magnusii mitochondria is primarily mediated by a sodiumindependent pathway. This could be a passive Ca 2ϩ /nH ϩ antiporter, with ⌬pH as a driving force, as in liver mitochondria. Whether indeed these transporters are related at the molecular level requires further studies.
The data presented demonstrate, to our knowledge for the first time, that mitochondria from a yeast species are endowed with independent systems for influx and efflux of Ca 2ϩ . Fig. 8   FIG. 8 summarizes the modulation of Ca 2ϩ uptake and efflux pathways in E. magnusii mitochondria and their possible physiological implications. An increase in the cytosolic Ca 2ϩ concentration would activate Ca 2ϩ uptake by respiring mitochondria via the Ca 2ϩ uniporter, with ⌬⌿ generated across the inner mitochondrial membrane as the driving force (7). This would establish a higher steady-state concentration of free Ca 2ϩ required for the activation of Ca 2ϩ -sensitive matrix enzymes, particularly of Ca 2ϩ -sensitive, NAD ϩ -dependent dehydrogenases supplying reducing equivalents to the respiratory chain (3). In our preliminary experiments, we found that the pyruvate dehydrogenase complex in E. magnusii mitochondria was activated by submicromolar [Ca 2ϩ ]. The Ca 2ϩ that accumulated in the matrix would be buffered by binding to P i and other substances. When Ca 2ϩ uptake exceeds the matrix Ca 2ϩ -binding capacity, as in the absence of added P i , an excessive rise in matrix [Ca 2ϩ ] would trigger Ca 2ϩ efflux primarily by a sodiumindependent mechanism, i.e. electroneutral Ca 2ϩ /2H ϩ antiport, or by activation of a Ca 2ϩ -specific channel (7). The result would be an increase in medium [Ca 2ϩ ], but the ensuing increased rate of Ca 2ϩ uptake in the Ca 2ϩ cycling would not be enough to lower ⌬⌿ (Fig. 2).
In conclusion, these data demonstrate that E. magnusii mitochondria have a specific Ca 2ϩ efflux pathway that is activated by elevated matrix free [Ca 2ϩ ]. In many respects, it is similar to the sodium-independent Ca 2ϩ efflux pathway in mitochondria from non-excitable mammalian cells. Thus, in both types of mitochondria, mitochondrial Ca 2ϩ efflux is modulated by the same substances. The efflux in E. magnusii was spontaneous and not due a decline in ⌬⌿ or induction of MPT.