Modulation of oxidative phosphorylation by Mg2+ in rat heart mitochondria.

The effect of varying the Mg2+ concentration on the 2-oxoglutarate dehydrogenase (2-OGDH) activity and the rate of oxidative phosphorylation of rat heart mitochondria was studied. The ionophore A23187 was used to modify the mitochondrial free Mg2+ concentration. Half-maximal stimulation (K0.5) of ATP synthesis by Mg2+ was obtained with 0.13 +/- 0.02 mM (n = 7) with succinate (+rotenone) and 0.48 +/- 0.13 mM (n = 6) with 2-oxoglutarate (2-OG) as substrates. Similar K0.5 values were found for NAD(P)H formation, generation of membrane potential, and state 4 respiration with 2-OG. In the presence of ADP, an increase in Pi concentration promoted a decrease in the K0.5 values of ATP synthesis, membrane potential formation and state 4 respiration for Mg2+ with 2-OG, but not with succinate. These results indicate that 2-OGDH is the main step of oxidative phosphorylation modulated by Mg2+ when 2-OG is the oxidizable substrate; with succinate, the ATP synthase is the Mg2+-sensitive step. Replacement of Pi by acetate, which promotes changes on intramitochondrial pH abolished Mg2+ activation of 2-OGDH. Thus, the modulation of the 2-OGDH activity by Mg2+ has an essential requirement for Pi (and ADP) in intact mitochondria which is not associated to variations in matrix pH.

The notion that the cytosolic concentration of free Mg 2ϩ ([Mg 2ϩ ] c ) 1 had a constant value around 1 mM under different conditions has changed in recent years. By using permeant fluorescent dyes and nuclear magnetic resonance, tissue-dependent variations in [Mg 2ϩ ] c in the range 0.4 of 0.8 mM have been observed, in response to several hormones and agonists. For instance, norepinephrine induced a net release of cellular Mg 2ϩ (1), while vasopressin induced Mg 2ϩ accumulation in isolated hepatocytes (2). Likewise, increments of 50% in [Mg 2ϩ ] c has been determined, after stimulation with the muscarinic agonist carbachol, and a 10% increase was observed, after addition of forskolin in rat sublingual mucous acini (3). In acinar pancreatic cells, addition of acetylcholine or cholecystokinin-octapeptide promoted a significant diminution in [Mg 2ϩ ] c (4). Arginine-vasopressin and endothelin-1 induced an incre-ment in [Mg 2ϩ ] c in muscle cells, probably through a Ca 2ϩmediated mechanism (5). Depletion of inositol 1,4,5-trisphosphate-sensitive Ca 2ϩ stores, induced by ␣-adrenergic agonists, activated the uptake of Mg 2ϩ by these organelles (6).
Extracellular ATP stimulated the release of 40% of cellular Mg 2ϩ in ascites cells (7); it was proposed that cAMP promoted Mg 2ϩ release through the activation of a plasma membrane Na ϩ /Mg 2ϩ antiporter (8). However, a report indicating that addition of cAMP also induced a net release of 20 -25% of total Mg 2ϩ in rat liver mitochondria (9) was not confirmed (10). In beef heart mitochondria, the transition from basal (state 4) to active (state 3) respiration led to a small, but significant elevation in the mitochondrial matrix free Mg 2ϩ concentration ([Mg 2ϩ ] m ) from 0.5 mM to 0.6 -0.7 mM. This increase in [Mg 2ϩ ] m persisted during ATP synthesis, until added ADP was exhausted; at this time [Mg 2ϩ ] m returned to basal levels. These variations in [Mg 2ϩ ] m were inhibited by oligomycin (11). An elevation in [Mg 2ϩ ] m from 44 M to 1.69 mM also induced a stimulation in the rate of citrulline synthesis in rat liver mitochondria (12). Modulation of mitochondrial glutaminase by 0 -2 mM Mg 2ϩ has also been observed (13).
All of these reports describing an active movement of Mg 2ϩ in cells and mitochondria of different tissues and in response to different agonists suggest that Mg 2ϩ may play a role as a second messenger in the cell. In this work, we show that variations of external Mg 2ϩ , and hence in [Mg 2ϩ ] m , can modulate the activities of the 2-OGDH and the ATP synthase and, in consequence, Mg 2ϩ may affect the rate of oxidative phosphorylation in isolated rat heart mitochondria.

MATERIALS AND METHODS
Rat heart mitochondria were isolated from male Wistar rats of 250 -300-g weight according to a previously described method using the protease type XXVII (Nagarse) from Sigma (14).
Dye Loading-Heart mitochondria were loaded with Mag-Fura-2 or BCECF (Molecular Probes) by incubating 30 -40 mg of mitochondrial protein in 2 ml of a medium composed of 250 mM sucrose, 10 mM HEPES, 1 mM EGTA (SHE medium), 1 mM MgCl 2 , 1 mM ADP, 0.2% fatty acid-free bovine serum albumin, pH 7.4, and 5 M Mag-Fura-2/AM or BCECF/AM at 25°C for 20 min. At the end of this incubation period, mitochondria were diluted 10 -15 times with ice-cold SHE medium ϩ 0.2% bovine serum albumin, centrifuged, resuspended in 1 ml of SHE medium, and kept on ice until use. Mitochondria loaded by following this procedure showed higher respiratory control values than nonloaded mitochondria, 8 Mg 2ϩ present in the incubation medium; 0.005% (v/v) Triton X-100 was added to ensure complete Mg 2ϩ equilibration across the membrane. R max was obtained after further addition of 70 mM MgCl 2 . Calculation of [Mg 2ϩ ] m was made using the following equation (15): where K d(Mg2ϩ) is the dissociation constant for the Mg-dye complex in the mitochondrial matrix and S f and S b are the dye fluorescence intensities at 398 nm with zero and excess Mg 2ϩ , respectively. The K d(Mg2ϩ) value was determined experimentally to be 1.52 Ϯ 0.18 mM (n ϭ 5).
pH Determination-BCECF-loaded mitochondria (0.5 mg of protein/ ml) were incubated in KME medium containing 0.5 mM 2-oxoglutarate, 10 mM NaCl, 600 M ADP, 3.5 M oligomycin, 800 pmol of A23187/mg of protein and different concentrations of Mg 2ϩ , P i , or acetate. For pH calculations, a calibration plot was generated incubating 0.5 mg of protein/ml in the medium mentioned above, at the desired pH, in the presence of 2 M carbonyl cyanide m-chlorophenylhydrazone, 200 pmol of nigericin/mg of protein and 0.005% Triton X-100 to equilibrate all ion gradients. Excitation wavelengths were 450 and 500 nm; fluorescence was collected at 530 nm. The plot of pH values versus fluorescence ratio signal gives a straight line between pH 6.8 and 7.8.
ATP Synthesis-Mitochondria (1 mg of protein/ml) were incubated in KME medium containing 0.5 mM 2-oxoglutarate or 5 mM succinate (ϩ1 M rotenone), 10 mM NaCl, 10 mM glucose, 30 units of hexokinase, and 5 mM 32 P i (specific activity, 1-1.5 ϫ 10 6 cpm/ml, Cerenkov radiation), at 30°C. After 5 min, 1.2 mM ADP was added, and the reaction was stopped 30 s later by addition of 200 l of 30% (w/v) cold trichloroacetic acid. Excess 32 P i was extracted as described previously using acetone ϩ n-butyl acetate as organic solvents (16). Radioactivity of an aliquot of the aqueous phase was determined as 32 P i Cerenkov radiation in a scintillation counter.
Activity of 2-OGDH-Mitochondria (1 mg of protein/ml) were suspended in KME medium containing 1 mM 2-oxoglutarate, 10 mM NaCl, 600 M ADP, pH 7.25, and different concentrations of Mg 2ϩ and P i at 30°C. Matrix NAD(P)H formed was determined following mitochondrial intrinsic fluorescence at 460 nm with the excitation wavelength at 340 nm. To obtain the fluorescence minimum, mitochondria were incubated in the absence of added substrates until endogenous substrates were depleted (approximately 5-8 min) (NAD(P)H ϭ 0%); the fluorescence maximum was reached by adding 5 M rotenone for complete reduction of NAD(P) ϩ (NAD(P)H ϭ 100%) at the end of each experiment.
The membrane potential was also quantitatively measured using the distribution of [ 3 H]TPP ϩ . Mitochondria (1.5 mg protein/ml) were suspended in 500 l of KME medium containing 5 mM P i , 10 mM NaCl, 0.8 M [ 3 H]TPP (specific activity, 4 -5 ϫ 10 4 cpm/ml) at 30°C and different concentrations of Mg 2ϩ . After 5 min, 800 pmol of A23187/mg of protein were added; 3 min later 1 mM 2-oxoglutarate was added, and the incubation was continued for another 3 min. Then, mitochondria were centrifuged at 14,000 rpm for 1 min in a microcentrifuge. Aliquots from the pellet and supernatant were taken to measure the [ 3 H]TPP ϩ distribution; the membrane potential was determined as described previously (19).
Oximetry Assays-Mitochondrial respiration was measured using an oxygen Clark-type electrode. Mitochondria (0.6 mg of protein/ml) were incubated in KME medium containing 1 mM 2-oxoglutarate, 10 mM NaCl, 1 or 5 mM P i , and 800 pmol of A23187/mg of protein. After 5 min, 600 M ADP was added, and the change in the rate of respiration was measured.
Matrix ATP and ADP Content-Mitochondria (2.5 mg of protein/ml) were incubated in KME medium plus 5 mM succinate and 2 M rotenone at 30°C for 10 min under orbital shaking. Then, 3% (v/v) cold perchloric acid, 25 mM EDTA was added, the suspension was centrifuged, and the supernatant neutralized for enzymatic determination of ATP and ADP. Essentially identical results were obtained when mitochondria were previously sedimented in a microcentrifuge at 6 -10°C and further denaturalized by the addition of perchloric acid.

RESULTS
The increase in the external Mg 2ϩ concentration induced a proportional, but small elevation in [Mg 2ϩ ] m in rat heart mitochondria (Fig. 1). This Mg 2ϩ gradient ([Mg 2ϩ ] m /[Mg 2ϩ ] ex ) showed a slope of 0.066, in the range 0 -3 mM externally added Mg 2ϩ , indicating that Mg 2ϩ does not easily equilibrate across the mitochondrial inner membrane, probably due to a slow Mg 2ϩ influx, or to an active Mg 2ϩ efflux. Similar results were previously reported for rat liver mitochondria (12). To accelerate the equilibration of Mg 2ϩ , the divalent cation ionophore A23187 was added. Fig. 1 shows that the ionophore modifies the steady-state concentration of matrix Mg 2ϩ , although equil- ibration with external Mg 2ϩ was not complete. Since mitochondria incubated with A23187 conserved the H ϩ gradient, it was not unexpected that A23187 did not produce complete equilibration of Mg 2ϩ concentrations across the mitochondrial membrane. Moreover, A23187 seems to be a weak Mg 2ϩ ionophore due to a low affinity and poor mobility across the mitochondrial membrane (20 -22). Nonetheless, the addition of A23187 allowed a more rapid manipulation of matrix Mg 2ϩ in a lower range of external Mg 2ϩ concentrations. The use of 1600 pmol of 4-bromo-A23187/mg of protein, instead of A23187, resulted in a curve very similar to that shown in Fig. 1 in the absence of ionophore (data not shown).
The addition of A23187 to mitochondria incubated in the absence of added MgCl 2 decreased [Mg 2ϩ ] m from 0.49 to 0.02 mM and induced a significant diminution of matrix ATP/ADP ratio and the ADP ϩ ATP content (see Fig. 1, inset). The further addition of 3 mM Mg 2ϩ , in the presence of A23187, increased [Mg 2ϩ ] m from 0.02 to 1.15 mM and preserved matrix ATP/ADP ratio and ADP ϩ ATP content at high values. Although a correlation between [Mg 2ϩ ] m and the ATP/ADP ratio (or ATP content) was not found for mitochondria incubated with 3 mM Mg 2ϩ and with or without A23187, it is apparent from the data of Fig. 1 that, in the presence of A23187, the addition of external Mg 2ϩ modified both the [Mg 2ϩ ] m and the ATP/ADP ratio, which may affect the rate of oxidative phosphorylation.
The rate of oxidative phosphorylation, assayed in the presence of A23187, depended on Mg 2ϩ concentration in the incubation medium (Fig. 2). Since the sensitivity to Mg 2ϩ depended on whether succinate (ϩrotenone) or 2-oxoglutarate (2-OG) was used (p Ͻ 0.05), the data of Fig. 2 suggest the existence of at least two sites of modulation by Mg 2ϩ . These sites may be located in the phosphorylating system (i.e. the ATP synthase or the adenine nucleotide translocase) during succinate oxidation, and at the level of 2-OGDH for 2-OG oxidation. Replacement of Mg 2ϩ by Mn 2ϩ also induced an activation of ATP synthesis, but at higher concentrations (K 0.5 values for Mn 2ϩ were 0.60 Ϯ 0.047 mM (n ϭ 3) with succinate and 0.92 Ϯ 0.052 mM (n ϭ 3) with 2-OG as a substrate).
Since succinyl-CoA synthase also requires Mg 2ϩ , its contribution to the uptake of 32 P i was assayed. In the inset of Fig. 2, it is shown that substrate level phosphorylation by the Krebs cycle accounted for up to 40 -50% of total ATP synthesis during oxidative phosphorylation with 2-OG as an oxidizable substrate. As substrate-level phosphorylation and oxidative phosphorylation with 2-OG showed different sensitivities to Mg 2ϩ , an effect of Mg 2ϩ on sites different from succinyl-CoA synthase seemed likely.
To discard the participation of contaminating ATPases, in the ATP synthesis assays, hexokinase ϩ glucose was used to capture ATP generated by oxidative phosphorylation. This prompted us to determine the Mg 2ϩ dependence of hexokinase. Under the conditions of ATP synthesis (see Fig. 2), K 0.5 values of hexokinase for ATP-Mg were 92 M, in the presence of 400 M ATP, and 31 M in the presence of 100 M ATP. These two concentrations of added ATP represent the maximal level of ATP synthesis (for 1 mg of protein/ml in 30 s at 30°C) during oxidative phosphorylation with succinate and 2-OG, respectively. The sensitivity of hexokinase to Mg 2ϩ revealed that this enzyme is not involved in the lower sensitivity of oxidative phosphorylation to Mg 2ϩ with 2-OG as a substrate (see Fig. 2). However, in the presence of succinate, the sensitivity of oxidative phosphorylation to Mg 2ϩ might result from a mixed response of both hexokinase and ATP synthase to Mg 2ϩ . However, an essentially identical sensitivity of oxidative phosphorylation to Mg 2ϩ was observed in the absence of hex- okinase (K 0.5 ϭ 0.13 Ϯ 0.014 mM, n ϭ 3), with succinate (ϩrotenone).
The change in the magnitude of the membrane potential, as estimated from the distribution of TPP ϩ , was initially used to monitor indirectly variations in 2-OGDH activity in mitochondria incubated with limiting concentrations of 2-OG (Fig. 3A). A membrane potential (⌬) of 130 mV in the absence of added ADP and Mg 2ϩ and in the presence of A23187 and P i , was determined. This value increased to 140 mV by the increase in [Mg 2ϩ ] ex (Fig. 3A, circles). With 600 M ADP, steady state ⌬ diminished to 112 mV by increasing [Mg 2ϩ ] ex (Fig. 3A, squares), due to stimulation of ATP synthesis. Although under these last conditions the oxidative system was activated by Mg 2ϩ , the diminution in ⌬ indicated that, at Mg 2ϩ concentrations of 0 -1.5 mM, activation of the phosphorylating system of the pathway by Mg 2ϩ was predominant. At higher Mg 2ϩ concentrations (Ͼ1.5 mM), activation of the oxidative system prevailed over that of the phosphorylating system, resulting in ⌬ values larger than those obtained at zero Mg 2ϩ (data not shown).
The enhancement of ⌬ up to 162 mV by increasing [Mg 2ϩ ] ex , in the presence of ADP ϩ P i ϩ oligomycin (Fig. 3A, triangles), which was larger than that reached at the same concentration of [Mg 2ϩ ] ex in the absence of ADP, indicated that ADP was a modulator of the Mg 2ϩ activation. In the absence of added Mg 2ϩ , removal of P i markedly diminished ⌬ (Fig. 3A, diamonds). In the absence of P i , ⌬ increased when [Mg 2ϩ ] ex was increased, but only to 105 mV. This latter observation prompted us to determine the effect of different concentrations of P i on the activity of 2-OGDH.
In comparison to the [ 3 H]TPP ϩ method, the absorbance difference of safranin O (Fig. 3B) allows for continuous monitoring of ⌬, and a large number of experiments with the same mitochondrial preparation. Using the safranin O signal, the increase in P i concentration (in the presence of ADP ϩ oligomycin) potentiated the activating effect of Mg 2ϩ on the steady state value of ⌬ (Fig. 3B). Thus, the half-maximal stimulation of ⌬ by Mg 2ϩ was decreased and the maximal value of ⌬ was elevated by increasing P i concentrations. This effect of P i was not apparent in the absence of added ADP. Similar results to those of Fig. 3B were obtained by measuring the [ 3 H]TPP ϩ distribution under the same conditions (data not shown).
The activity of 2-OGDH was also measured, following the level of reduction of matrix pyridine nucleotides. In the absence of added P i , the increase in [Mg 2ϩ ] ex did not promote the generation of NAD(P)H (Fig. 4A). However, an increase in P i concentration induced both a decrease in the K 0.5 value for Mg 2ϩ and an increase in the level of NAD(P)H reduction. The rate of respiration measured in the presence of ADP ϩ oligomycin (state 4) was also stimulated by increasing [Mg 2ϩ ] ex (Fig.  4B). Again, the presence of increasing P i concentrations potentiated the stimulation by Mg 2ϩ , through a diminution in the K 0.5 value for Mg 2ϩ and an increase in the maximal rate of respiration. Thus, the data of Figs. 3 and 4 indicate that P i , in the presence of ADP, potentiates the activating effect of Mg 2ϩ on 2-OGDH activity.
The study of the effect of different P i concentrations on the Mg 2ϩ sensitivity of ATP synthesis and state 3 respiration supported by succinate (ϩrotenone) revealed a negligible effect on the K 0.5 value for Mg 2ϩ , indicating that the effect of P i was exerted only at the Krebs cycle level. Lack of Mg 2ϩ activation on state 4 respiration with succinate (ϩrotenone) as substrate and oligomycin (data not shown), discarded the possibility that Mg 2ϩ activated the respiratory chain.
Matrix acidification brought about by the P i uptake might be involved in Mg 2ϩ activation of 2-OGDH. The activating effect of 5 mM P i on the stimulation of matrix NAD(P)H formation by Mg 2ϩ was not reproduced by addition of 10 or 20 mM acetate (data not shown); the final steady-state pH values in BCECFloaded mitochondria incubated with 10 mM acetate or 5 mM P i , in the presence of ADP, oligomycin, and A23187, were 6.91 and 6.88 with no added Mg 2ϩ , and 7.11 and 7.16 with 1 mM Mg 2ϩ , respectively. These results indicate that matrix acidification is not the mechanism involved in the P i potentiating effect.

DISCUSSION
The different sensitivity to Mg 2ϩ of the rate of oxidative phosphorylation, with 2-OG or succinate, suggested that different sites in the pathway are involved in the interaction with Mg 2ϩ . With the former substrate, the data on ⌬, matrix NAD(P)H, and respiratory rates indicated that the main site of interaction with Mg 2ϩ was 2-OGDH.
Since ADP-Mg is the true substrate for the ATP synthase (24), increasing [Mg 2ϩ ] m is expected to activate this enzyme. However, the substrates for adenine nucleotide translocase are free ADP and ATP (25). During ATP synthesis, external ADP exchanges with matrix ATP. As ATP has a higher affinity for Mg 2ϩ than ADP, increasing [Mg 2ϩ ] m is expected to inhibit the adenine nucleotide translocase activity through the diminution of the internal substrate. Therefore, the effect of increasing [Mg 2ϩ ] m on the rate of oxidative phosphorylation is not readily apparent. As the elevation of external Mg 2ϩ , and hence [Mg 2ϩ ] m , resulted in higher rates of ATP synthesis with succinate as substrate, it can be assumed that Mg 2ϩ activation of the ATP synthase prevailed over Mg 2ϩ inhibition of the adenine nucleotide translocase. Stimulation of the rate of oxidative arsenylation, an analogous process to oxidative phosphorylation, but without the participation of adenine nucleotide translocase (16), by Mg 2ϩ using succinate (data not shown), supported the interpretation of an activating effect of Mg 2ϩ on the ATP synthase.
In addition to a direct interaction of Mg 2ϩ with the oxidative phosphorylation enzymes, Mg 2ϩ might also perturb matrix Ca 2ϩ homeostasis, and hence, affect the rate of ATP synthesis (16,19,26) (reviewed in Moreno-Sá nchez and Torres-Má rquez (27)). For instance, Mg 2ϩ might compete with Ca 2ϩ for the same binding sites in 2-OGDH. A decreased Ca 2ϩ sensitivity by increasing Mg 2ϩ has been observed for the NAD ϩ -isocitrate dehydrogenase (28), whereas an enhanced Ca 2ϩ sensitivity was described for the pyruvate dehydrogenase phosphatase (29). Although the sensitivity of 2-OGDH to Ca 2ϩ , at different Mg 2ϩ concentrations, has not yet been determined, Panov and Scarpa (30) reported that 2-OGDH can be activated synergistically by both Mg 2ϩ and Ca 2ϩ , implying the existence of different binding sites.
Panov and Scarpa (30) also determined a dissociation constant (K d ) for Mg 2ϩ of 25 M in the isolated 2-OGDH, with saturating concentrations of thiamine pyrophosphate, coenzyme A, and NAD ϩ . Although such a K d value for Mg 2ϩ is lower than the K 0.5 value obtained in this study (0.48 mM, see Fig. 2), it can be argued that the matrix concentrations of the 2-OGDH coenzymes in intact heart mitochondria may be limiting, and that 2-OGDH activity is not the only controlling step of the pathway (25). The value of the K d or K 0.5 for Mg 2ϩ may establish the physiological relevance of variations in [Mg 2ϩ ] m . Thus, a K 0.5 value of 0.48 mM would appear as more physiologically relevant for modulating 2-OGDH activity and the rate of oxidative phosphorylation, since this concentration is in the range of [Mg 2ϩ ] m in intact mitochondria (31,32).
It should be noted, however, that the estimated K 0.5 values for Mg 2ϩ refers to the external Mg 2ϩ concentrations, which were not fully equilibrated with the mitochondrial matrix by A23187 (cf. Fig. 1). Thus, the K 0.5 value of 0.48 mM for external Mg 2ϩ corresponds to a [Mg 2ϩ ] m of 140 M, which is slightly below the physiological range. Higher K 0.5 values for Mg 2ϩ were determined at low P i concentrations. For instance, a K 0.5 value of 1 mM for Mg 2ϩ was observed in NAD(P)H formation with 1 mM P i (see Fig. 4A). Such a K 0.5 value was diminished to 0.5 mM by increasing P i concentration up to 3 mM P i . The corresponding [Mg 2ϩ ] m for 1 mM external Mg 2ϩ would be 350 M, a value well within the physiological range. A variation in the cytosolic P i concentration from 0.83 to 3.1 mM induced by epinephrine was established in rat heart (33). Therefore, physiological modulation of the 2-OGDH activity by Mg 2ϩ may depend on the level of cytosolic (and matrix) P i .
Other possible sites of modulation by Mg 2ϩ during oxidative phosphorylation supported by 2-OG oxidation were the succinyl-CoA synthase, the ATP synthase and hexokinase (in the experiments of 32 P i incorporation into ATP). However, the Mg 2ϩ sensitivity of these three enzymes showed that their saturation by Mg 2ϩ was fully achieved at concentrations (Ͻ0.2-0.3 mM) that stimulated oxidative phosphorylation by less than 40%. The lack of stimulation of state 4 respiration by Mg 2ϩ in mitochondria that oxidized succinate, in the presence of oligomycin, discarded an effect of Mg 2ϩ at the level of the respiratory chain. Thus, these results indicate that 2-OGDH is one (but not the only) of the main controlling steps of oxidative phosphorylation (see also Moreno-Sá nchez et al. (26)), at nonsaturing Mg 2ϩ concentrations. In this respect, control of the rate of oxidative phosphorylation by changes in the spermine/ Mg 2ϩ rates, without a concomitant increase in [Ca 2ϩ ] m , has been shown in dog pancreas mitochondria (19).
Modulation of the 2-OGDH activity by adenine nucleotides is well established (23). A synergistic effect by Ca 2ϩ and adenine nucleotides has been described (28). Mg 2ϩ also activates 2-OGDH (30) (this work), but in contrast to other enzyme effectors, the mechanism of action is by enhancing the catalytic enzyme capacity (k cat ), rather than by increasing substrate affinity. Potentiation of the modulating effect of Mg 2ϩ by P i , although clearly demonstrated in this work, is somewhat puzzling. There is a report describing an activation of purified 2-OGDH by a high concentration of P i (Ͼ10 mM), through the diminution of the K 0.5 for 2-OG (34). Moreover, the P i potentiating effect could be through promoting changes in matrix pH, since modulation of 2-OGDH activity by pH has also been reported (35). However, substitution of acetate for P i , to induce similar matrix pH values, did not restitute the Mg 2ϩ sensitivity of 2-OGDH. Thus, a direct interaction of P i with the enzyme is likely to occur. From the present findings, the question that arises is to what extent and how P i and Mg 2ϩ affect the interplay of the other well described effectors, such as Ca 2ϩ and adenine nucleotides, and the coenzymes NAD ϩ , thiamine pyrophosphate, and coenzyme A, on 2-OGDH activity.