Characteristics of the Movement of I<+ across the Mitochondrial Membrane and the Inhibitory Action of Tl+

The incubation of mitochondria in mixtures that contain phosphate, NaCl, oxidizable substrate, and ethylenediaminetetraacetate induces the efflux of K-+. This process depends on electron transport and on the cyclic movement of phosphate across the membrane. Sodium ions, Li-+, or Cs-+ to a smaller extent, are required for maximal release of K-+. Potassium ions do not induce net efflux of internal K-+, but instead prevent the Na-+-induced release of K-+. Significant K-+ influx takes place in K-+-depleted mitochondria through a process with characteristics which are almost identical with those in which K-+ release takes place. As Na-+ inhibits the uptake of K-+, it is suggested that the movement of K-+ across the membrane is controlled by the cationic environment. Thallous ion, at concentrations that do not affect oxidative phosphorylation, was found to be an effective inhibitor of the influx and the efflux of K-+. The inhibitory effect of Tl-+ seems to be specific for K-+ since it does not affect the movement of Na-+. Mitochondria bind 10 to 15 nmol of 204-Tl-+ per mg of protein through an energy-independent process.


SUMMARY
The incubation of mitochondria in mixtures that contain phosphate, NaCl, oxidizable substrate, and ethylenediaminetetraacetate induces the efflux of K+. This process depends on electron transport and on the cyclic movement of phosphate across the membrane. Sodium ions, Lit, or Cs+ to a smaller extent, are required for maximal release of I(+. Potassium ions do not induce net efflux of internal K+, but instead prevent the Nat-induced release of K+. Significant Kf influx takes place in K+-depleted mitochondria through a process with characteristics which are almost identical with those in which Kf release takes place. As Na+ inhibits the uptake of K+, it is suggested that the movement of K+ across the membrane is controlled by the cationic environment. Thallous ion, at concentrations that do not affect oxidative phosphorylation, was found to be an effective inhibitor of the influx and the efllux of K+. The inhibitory effect of Tl+ seems to be specific for K+ since it does not affect the movement of Na +. Mitochondria bind 10 to 15 nmol of 204Tlf per mg of protein through an energy-independent process. Although ionophore-mediated translocation of monovalent cations across the mitochondrial membrane has been extensively studied (for review see Refs. 1 and 2), little work has been done on the intrinsic capacity of the mitochondria to move K+ across its inner membrane. Perhaps this is due to the finding that mitochondria maintain their K+ against a concentration gradient (3), which suggested that the membrane was impermeable to K+. lMoreover, mitochondria retain K+ in conditions in which extensive loss of Na+ takes place (4). However, in relatively long periods of incubation at 38", mitochondria released a significant amount of K+ that could be taken up again provided an energy source was introduced (5,6) ; also Gamble (7) found that digitonin particles possess an energy-dependent K+ transport. It has also been observed that EDTA increases the permeability of the mitochondria to monovalent cations (8-lo), and more recently, Gamble (11) found that citrate could mediate the uptake of K+. Therefore, these findings suggest that K+ movements in mitochondria are subject to some type of control. This is of importance since monovalent cations (12-16) and their movement (17-20) across the membrane affect oxidative phosphorylation and related reactions. In this work we describe some characteristics of the energy-and cation-dependent translocation of Kf in mitochondria.
Also the inhibitory effect of thallous ion on the movement of Kf is described. A possible inhibitory effect of Tlf on K+ translocation was studied because its atomic radius is similar to that of K+ (1.40 and 1.33 A, respectively), but it binds more strongly to proteins (2 1, 2 2).

MATERIALS AND METHODS
Rat liver mitochondria were prepared as described elsewhere (14) in 0.25 M sucrose plus 1 mM EDTA adjusted to pH 7.3 with Tris base. In some cases mitochondria were prepared in the absence of EDTA. The release of K+ was measured as described before (15). Our isolated intact mitochondria had 100 to 130 nmol of KC per mg of protein. Experiments with preparations with less than 100 nmol were discarded. In some experiments inorganic phosphate was determined in the 6% trichloroacetic extract of a mitochondrial suspension according to Sumner (23). Uptake of K+ was measured by incubating mitochondria depleted of K+ for 3 min in mixtures of variable composition (see under "Results"); subsequently the mixture was diluted with sucrose with or without EDTA and centrifuged at 10,000 X g for 5 min. The mitochondria were washed twice with sucrose or sucrose plus EDTA. The K+ content of these mitochondria was measured by flame photometry as described elsewhere (15).
Oxygen uptake was measured polarographically (Yellow Springs Instrument Co.), and ADP to oxygen ratios were measured according to Estabrook (24).
The binding of zo4T1+ (Amersham/Searle) to mitochondria was determined by incubating the particles in mixtures that contained various concentrations of *OaTl+ and thereafter filtered through Millipore.
The filter was washed with 2.0 ml of 0.25 M sucrose and transferred to a scintillation vial that contained 10 ml of Bray's solution. The samples were counted with a Packard scintillator with a '4C window. As the Millipore filters bound z"aT1+, a correction was made by subtracting the radioactivity found in filters through which aliquots of the reaction mixture (without mitochondria) had been passed and washed with 2.0 ml of sucrose. Protein was determined according to Lowry et al. (25).

RESULTS
Characteristics of Movement of Kf across Membrane-Although the uptake and release of K+ by mitochondria were measured independently and under different conditions the results are presented simultaneously to facilitate a comparison of the characteristics of the two processes. Maximum efflux and influx of K+ require phosphate and EDTA (Table I). Outward and inward translocation of Kf also require electron transport, as evidenced by the sensitivity of the two processes to rotenone (26).   Table  I except that the  complete  mixture  (see Table  I (Table IV).
An attempt to study the role of phosphate on K+ movements was made by studying the effect of mersalyl on K+ efflux. The concentration of mersalyl used was that known to block the phosphate carrier (27,28). The change in the concentration of phosphate in mitochondria was measured simultaneously with the amount of K+ lost (Table V). Mersalyl inhibited the efflux of IF by more than 50% which indicates that the movement of phosphate across the membrane is required for maximal release. Moreover, it may also be concluded that phosphate undergoes a cyclic movement across the membrane during K+ release since intramitochondrial phosphate does not change significantly. It should be pointed out that we have observed significant variability in the uptake of K+. In 36 experiments that have been carried out in strictly identical conditions with 50 m&f K+ in the uptake mixture, it has been observed that two preparations of prevented K+ movements at concentrations lower than EDTA (Table II).
An important difference between K+ influx and efflux is that Na+ is required for maximal release of K+ (Table I). Therefore, the effect of various cations and changes in the concentration of Na+ was studied on Kf efflux. Sodium ions induced maximal K+ efflux at 40 mM. The results showed that at 8 mM Tl+ the leakage of Kf is reduced by about 50% (Fig. 1A) and that the effect of Tl+ occurs at all of the concentrations of Na+ employed (Fig. 1B) which suggests that Na+ does not compete with TV.
It is possible that Tl+ affected the loss of Kf by inhibiting the action of Na+, by blocking the pathway through which K+ is released, or by acting on both systems. However, Tl+ at a concentration of 5 mM did not affect the previously described (14, 16) Na+-induced release of oxygen uptake (Fig. 2). Apparently Tlf prevents the movement of K+, but not that of Na+.
The effect of Tl+ on the uptake of K+ by K+-depleted mitochondria was also studied. Tl+ inhibits the uptake of K+, but the results shown in Fig. 3 indicate that the extent of the inhibition depends on the relative concentrations of K+ and TV. The inhibitory action of Tl+ is more marked at the higher Tl+:K+ ratios, which suggests that a competition exists between the two cations. Unfortunately, due to the nature of the experimental system employed a detailed kinetic analysis of the behavior of K+ and Tl+ cannot be made at the present.
The preceding experiments suggested that Tl+ acts on a system responsible for the inhibition of K+ movements across the membrane. Thus the amount of Tl+ that binds to mitochondria was measured. The data of Fig. 4 show the amount of 204Tl+ that appears in mitochondria incubated in sucrose (+ rotenone) or in the regular Kf-depletion mixture. As in both systems almost identical results were obtained, the amount of Tlf that appears in mitochondria probably corresponds to "bound" Tl+ and not Tl+ that has accumulated. At 8 mM TV, the concentration that induces maximal inhibition of Kf movements, about 12 nmol of Tl+ per mg of protein are bound.
The effect of Tl+ on K+ movements could be secondary to its action on electron transport or on the phosphorylation reaction. However, this possibility seems unlikely since concentrations of  Table  I  Tl+ which suppress K+ movements do not affect the respiratory rates or ADP to oxygen ratios of mitochondria (Fig. 5).

DISCUSSION
The results of this work show that under optimal conditions and in a period of 3 min close to 100 nmol of Kf are released per mg of mitochondrial protein and that a significant, although variable, amount of K+ can move into the mitochondria. K+ movements of this magnitude have been observed only in the presence of ionophores (29) and uncouplers (30) and in prolonged periods of incubation (5, 6). Optimal movements of K+ take place in mixtures that contain EDTA and phosphate and in which electron transport occurs; in the case of Kf release, a monovalent cation other than K+ is also required. In the absence of either of these components, the movement of K+ is diminished, but not abolished. This suggests that electron transport, phosphate, EDTA, and the cationic environment contribute individually to poise the direction and the magnitude of K+ movements. As efflux SCHEME 1 and influx of Kf have similar characteristics except for their metal ion requirements, i.e. prevention of the Na+-induced efflux by K+ and inhibition of K+ uptake by Na+, it is logical to assume that the relative concentrations of Ea+ and K+ govern the direction of the flux of K+ across the membrane.
With respect to the events that occur during the movement of Kf, a mechanism is suggested in Scheme 1. Int.ernal OH-was generated during electron transport exchanges for external phosphate (31, 32). Once on the inside, phosphate would induce the influx of an external cation. In a second step, internal phosphate and an internal cation (K+) would be released from the mitochondria. This set of events would account for the observations that during the loss of K+ from the mitochondria, phosphate undergoes a cyclic movement across the membrane (Table V), and also of the requirement for electron transport.. n'evertheless, other alternatives are equally possible.
It has been reported (8-10) that EDTA reduces the Mg2+ content of the mitochondria (8) ; thus it has been suggested that bound Mg2+ is important for controlling the permeability of the membrane. However, K+ influx in mitochondria which had been depleted of K+ in the presence of EDTA also requires EDTA. Therefore, other factors besides the reported decrease in hlg'+ content, such as binding of EDTA to the membrane, should be responsible for the increase in the permeability of the membrane to monovalent cations. An objective of this work was to study whether or not movement of K+ is affected by Tl+. Tl+, because of its ability to bind more strongly than K+ to protein (21, 22), was considered a possible inhibitor of Kf movements. Moreover, Tl+ was recently reported to form complexes with valinomycin (33), the highly specific Kf ionophore (29). Also Tlf may substitute for K+ in the activation of the (iVa+-K+)-activated ATPase (34) and pyruvate kinase (22), and it is accumulated by erythrocytes The inhibition of the influx and efflux of Kf by Tlf indicates that both the inward and outward movements of K+ take place through the same molecular system. Apparently, the same molecular entity reacts with external and internal Kf. Since Tl+ does not inhibit the Na+-stimulated oxygen uptake, the system seems specific for I(+. As a corollary, it may be concluded that Na+ movements across the membrane occur through a system that is different from that through which Kf is translocated.
The amount of bound Tl+ required to inhibit K+ translocation is of the order of 10 to 15 nmol per mg of mitochondrial protein. Since in our experimental conditions nonspecific binding cannot be excluded, this value is probably higher than the actual concentration of entities responsible for K+ translocation. Moreover, we cannot discard the fact that phospholipids play a role in K+ translocation.