Anomalous oxygen-18 exchange during ATP synthesis in oxidative phosphorylation

The synthesis of ATP from highly enriched [18O]Pi by submitochondrial particles driven by succinate oxidation produces distributions of 18O-labeled ATP species that deviate from the distributions predicted by a simple model for the exchange. Control experiments indicate no change in isotopic distribution when [18O]ATP is synthesized from [18O]ADP by adenylate kinase, which is bound to the submitochondrial particles. The observed deviations are in the opposite direction from that produced by heterogeneity due to multiple pathways for ATP synthesis. Two types of complex models can account for the observed deviations. One model has nonequivalence of the Pi oxygens during the exchange reaction, due to incomplete randomization of the Pi oxygens during the reversible cycles of hydrolysis and synthesis of bound ATP. The other model assumes that, during each turnover, a slow transition must occur between a high-exchange and a low-exchange pathway.     

3-Deoxy-D-manno-octulosonate-8-phosphate synthase catalyzes the C-O bond cleavage of phosphoenolpyruvate

The mechanism of 3-deoxy-D-manno-octulosonate-8-phosphate (KDO8P) synthase was investigated. When [18O]-PEP specifically labeled in the enolic oxygen is a substrate for KDO8P synthase, the 18O is recovered in Pi. This indicates that the KDO8P synthase reaction proceeds with C-O bond cleavage of PEP similar to that observed in the 3-deoxy-D-arabino-heptulosonate-7-phosphate synthase catalyzed condensation of PEP and erythrose-4-phosphate (1). No evidence for a covalent enzyme-PEP intermediate could be obtained. No [32P]-Pi exchange into PEP nor scrambling of bridge 18O to non-bridging positions in [18O]-PEP was observed in the presence or absence of arabinose-5-phosphate or its analog ribose-5-phosphate. Bromopyruvate inactivated KDO8P synthase in a time dependent process. It is likely that bromopyruvate reacts with a functional group at the PEP binding site since PEP, but not arabinose-5-phosphate, protects against inactivation.     

A stereochemical method for detection of ATP terminal phosphate transfer in enzymatic reactions. Glutamine synthetase.

An isotope scrambling method is described for the detection of transient [Enz:ADP:P-X] formation from [18O]ATP in ATP-coupled enzyme reactions. The method makes use of torsional symmetry of the newly formed (see article) group in ADP. [18 O]ATP labeled in the betagama bridge oxygen was incubated with enzyme and reversible cleavage of the PbetaO -- Pgamma bond was detected by the appearance of 18O in the beta nonbridge oxygens of the ATP pool. Experiments with sheep brain and Escherichia coli glutamine synthetases show that cleavage of ATP of enzyme-bound ADP and P-X requires glutamate. The exchange catalyzed by the E. coli enzyme with glutamate occurs in the absence of ammonia and is partially inhibited by added NH4Cl, as expected if the exchange is in the mechanistic pathway for glutamine synthesis. The results provide kinetic support for a two-step mechanism where phosphoryl transfer from ATP to glutamate precedes reaction with ammonia.     

Investigations of the partial reactions catalyzed by pyruvate phosphate dikinase

The kinetic mechanism of pyruvate phosphate dikinase (PPDK) from Bacteroides symbiosus was investigated with several different kinetic diagnostics. Initial velocity patterns were intersecting for AMP/PPi and ATP/Pi substrate pairs and parallel for all other substrate pairs. PPDK was shown to catalyze [14C]pyruvate in equilibrium phosphoenolpyruvate (PEP) exchange in the absence of cosubstrates, [14C]AMP in equilibrium ATP exchange in the presence of Pi/PPi but not in their absence, and [32P]Pi in equilibrium PPi exchange in the presence of ATP/AMP but not in their absence. The enzyme was also shown, by using [alpha beta-18O, beta, beta-18O2]ATP and [beta gamma-18O, gamma, gamma, gamma-18O3]ATP and 31P NMR techniques, to catalyze exchange in ATP between the alpha beta-bridge oxygen and the alpha-P nonbridge oxygen and also between the beta gamma-bridge oxygen and the beta-P nonbridge oxygen. The exchanges were catalyzed by PPDK in the presence of Pi but not in its absence. These results were interpreted to support a bi(ATP,Pi) bi(AMP,PPi) uni(pyruvate) uni(PEP) mechanism. AMP and Pi binding order was examined by carrying out dead-end inhibition studies. The dead-end inhibitor adenosine 5'-monophosphorothioate (AMPS) was found to be competitive vs AMP, noncompetitive vs PPi, and uncompetitive vs PEP. The dead-end inhibitor imidodiphosphate (PNP) was found to be competitive vs PPi, uncompetitive vs AMP, and uncompetitive vs PEP. These results showed that AMP binds before PPi. The ATP and Pi binding order was studied by carrying out inhibition, positional isotope exchange, and alternate substrate studies.(ABSTRACT TRUNCATED AT 250 WORDS)     

Induction by nucleotide triphosphate hydrolysis of a form of sarcoplasmic reticulum ATPase capable of medium phosphate-oxygen exchange in presence of calcium.

Sarcoplasmic reticulum vesicles rendered leaky by exposure to alkaline pH, like intact vesicles, catalyze a rapid Mg2+-dependent exchange of oxygens of medium Pi with water. The exchange with 10 mM Pi is strongly inhibited by 0.15 mM Ca2+. Upon addition and hydrolysis of ITP or ATP, a rapid phosphate-oxygen exchange is observed even with 0.15 mM Ca2+ present and a definite but smaller exchange at 8 mM Ca2+. Oxygen exchange per Pi formed is greater with ITP than with ATP. When no Pi is initially present, the extent of oxygen exchange is increased with time of incubation as Pi is formed. With 18O-labeled Pi present, ATP hydrolysis accelerates 18O loss. The results show that much of the oxygen exchange occurs as a result of reversible binding of medium Pi. Thus the binding and cleavage of ITP or ATP overcomes the Ca2+ inhibition of the medium Pi in equilibrium HOH exchange. Such findings support the concept that the cleavage cycle includes a transient conformational form which can reversibly react with Pi to give a phosphoryl enzyme and resultant oxygen exchange or in a rate-limiting step decay to a form with high Ca2+ and NTP affinity.     

Two types of kinetic regulation of the activated ATPase in the chloroplast photophosphorylation system

The inhibition by light of chloroplast coupling factor ATPase is not due simply to competing photophosphorylation. This inhibition is only partially relieved by either an arsenate-pool trap for released phosphate, or a pyruvate kinase/phosphoenolpyruvate trap for ADP. Moreover, the amount of product return that does occur in the absence of trapping systems, ascertained by incorporation of 32Pi or [2-3H]ADP back into ATP during the hydrolysis reaction, is insufficient to account for the observed activity decrease. In intermediate pi:H2O oxygen exchange studies, the number of water oxygens incorporated into each molecule of Pi produced does not vary with light intensity during the ATPase assay. This indicates that the light-induced change in ATPase activity is not due to an alteration of rat constants involved in the forward and reverse partitioning of the E.ADP.Pi complex. In contrast, ammonium chloride, an uncoupler of photophosphorylation which stimulates membrane-bound coupling factor ATPase when added after light activation, causes a shift in the pattern of intermediate Pi:H2O oxygen exchange toward a lower number of water oxygens incorporated per Pi formed. The effect of NH4+ consistent with ATPase activity stimulation caused by enhanced partitioning forward of the E.products complex. These observations suggest the operation of two mechanisms of regulation of ATP ase activity during chloroplast de-energization. However, a direct effect of NH4+ on the coupling factor itself, independent of the membrane energization effect, cannot be ruled out by the present studies. Additional oxygen exchange experiments lead to the conclusion that the binding of ATP at a site catalyzing extensive ATP:H2O back exchange in the native chloroplast system ( Wimmer, M. J., and Rose, I. A. (1977) J. Biol. Chem. 252, 6769-6775) is different from the binding of ATP for net hydrolysis in the system activated for ATPase.     

Mechanism of oxygen exchange in actin-activated hydrolysis of adenosine triphosphate by myosin subfragment 1.

The gamma-phosphoryl groups of two intermediates (M-ATP and M-ADP-P1) in the pathway of MgATP hydrolysis by myosin undergo extensive oxygen exchange with water. Actin activates the overall rate of hydrolysis at a rate-limiting step which follows these exchange reactions. Thus, actin, by decreasing the turnover time of hydrolysis, would be expected to proportionately decrease the time available for oxygen exchange. Using subfragment 1 of myosin, the turnover time of hydrolysis can be varied over a wide range by changing the concentration of actin. An estimate for the rate constant of exchange can then be obtained by relating these turnover times to measured values for oxygen exchange (incorporation of 18O from H218O into the inorganic phosphate (Pi) released by hydrolysis). The results of such an experiment, with turnover times between 0.2 and 25 s, indicate that, for each gamma-phosphoryl group, one oxygen from the medium is added rapidly (to cleave the phosphoryl group or form a pentacoordinate phosphroyl complex); two more oxygens exchange with a rate constant, kc, of about 1 s-1; and a fourth oxygen exchanges slowly with ke about 0.2 s-1. The higher value is about 18 times smaller than the rate constant, 5-3, for the reverse cleavage step of the myosin pathway, which is postulated to be responsible for oxygen exchange. The data, then, indicate that the rate-limiting step for oxygen exchange is not k-3, but may be the rate of rotation of oxygens around the phosphorus atom, with one oxygen severely restricted by its binding to the active site. The finding that kc differs for the four oxygens in each phosphate group is related to past observations on myosin-catalyzed oxygen exchange.     

Purine nucleoside phosphorylase cleaves the C--O bond of ribose 1-phosphate. Evidence from the 18O shift in 31P NMR.

An equilibrium mixture of highly enriched [18(O)]Pi (represents the mixture of [[18(O)4]Pi, [[18(O)3]Pi, [18(O)2]Pi as represented in the figures, unless otherwise specified), alpha-D-ribose 1-[16(O)]phosphate, and hypoxanthine plus inosine was equilibrated with calf spleen purine-nucleoside phosphorylase (EC 2.4.2.1). The 31P NMR spectrum clearly indicated the formation of alpha-D-ribose 1-[18(O)4]-phosphate and of [16(O)]Pi. Incubation for the same time span in the absence of alpha-D-ribose 1-phosphate left the [18(O)4]Pi isotopic distribution unchanged. The results clearly demonstrated that the C--O bond of alpha-D-ribose 1-phosphate is cleaved in the enzymatic reaction. It is unlikely that the enzyme catalyzes the exchange of oxygen between Pi and H2O. Several possible mechanistic pathways are ruled out by the results, which demand attack by a phosphate oxygen at the anomeric C-1' atom.     

Stereochemical course of phosphokinases. The use of adenosine [gamma-(S)-16O,17O,18O]triphosphate and the mechanistic consequences for the reactions catalyzed by glycerol kinase, hexokinase, pyruvate kinase, and acetate kinase.

We report the synthesis of adenosine [gamma-(S)-16O,17O,18O]triphosphate, an isotopically labeled species of ATP that is chiral at the gamma-phosphoryl group, the configuration of which has been confirmed by independent stereochemical analysis. This molecule has been used as a substrate in the reactions catalyzed by glycerol kinase and by acetate kinase. The resulting samples of isotopically labeled sn-glycerol 3-phosphate and of acetyl phosphate have been used as substrates in the alkaline phosphatase mediated transfer of the chiral phosphoryl groups to (S)-propane-1,2-diol, whence the configuration at phosphorus has been determined [Abbott, S. J., Jones, S. R., Weinman, S. A., & Knowles, J. R. (1978) J. Am. Chem. Soc. 100, 2558]. It is shown that glycerol kinase and acetate kinase (and, by virtue of an earlier correlation, pyruvate kinase and hexokinase) proceed by pathways that result in inversion of the configuration at phosphorus. The sterochemical approach provides an access to the otherwise cryptic events that are involved in phosphoryl-group transfer within the ternary complexes of these kinases and their substrates.     

Characterization of phosphate oxygen exchange reactions catalyzed by myosin through measurement of the distribution of 18-O-labeled species

The change in the distribution of the phosphate species containing 0 to 4 18O oxygens per Pi was investigated during medium Pi equilibrium HOH exchange catalyzed by myosin subfragment 1. At 25 degrees C, a Pi molecule once bound loses an average of 3.9 of its original 4 oxygens prior to release which means that at least 100 reversals of the exchange reaction must have occurred. At 0 degrees C, only 3.4 of the 4 oxygens are lost prior to release indicating an average of 17 reversals. Distribution patterns are consistent with equivalent participation in the exchange reactions of all 4 oxygens of bound Pi. The intermediate exchange of Pi oxygens during hydrolysis of 18O-labeled ATP by myosin has also been investigated. The distribution of the product Pi species shows that there is an ATPase component in myosin preparations which hydrolyzes ATP without intermediate exchange. Presence of this component, which is likely a contaminating ATPase, provides a simple explanation of the apparent nonequivalence of phosphate oxygens which has been observed. When correction is made for this contaminant, characteristics of the myosin intermediate Pi equilibrium HOH exchange are similar to those of myosin subfragment 1 medium exchange, and intermediate exchange data are in much closer agreement with other kinetic measurements.     

Stereochemistry of phosphoryl transfer

A general method has been developed for the synthesis of chiral [16O,17O,18O]phosphate monoesters of known absolute configuration. An analytic method for determining the absolute configuration of chiral phosphate esters has also been developed, which is based on the isotope effects of 17O and 18O at phosphorus in the 31P nuclear magnetic resonance spectrum. These methods have shown that phosphoryl transfer catalysed by hexokinase, phosphofructokinase and pyruvate kinase occurs with inversion of configuration. This is most simply interpreted as an "in-line' transfer of the phosphoryl group between substrates in the enzyme-substrate ternary complex.     

Theoretical analysis of distribution of [18O]Pi species during exchange with water. Application to exchanges catalyzed by yeast inorganic pyrophosphatase

A theoretical analysis has been derived which allows the analytical calculation of the complete distribution of 18O-labeled Pi species expected to occur during medium Pi equilibrium HOH exchange of [18O]Pi and to be produced by intermediate Pi equilibrium HOH exchange during net hydrolysis of [18O]PPi or other labeled phosphate compounds. The observed distributions with catalysis by yeast inorganic pyrophosphatase are found to agree closely with the theoretical values indicating that the exchange reaction can be adequately described by a unique value of the partitioning of bound Pi between release from the enzyme versus formation of bound PPi with loss of an oxygen to the water. The limitations on the exclusion of other mechanisms are discussed. The extent of this partitioning does change, however, under some experimental conditions. At low pH, with activation by Mg2+ or Mn2+, the relative rate of release of Pi is found to increase. The extent of exchange is also dependent on the nature of the activating metal, being greatest with Co2+. During PPi hydrolysis with PPi in excess over Mg2+, a shift to lower extents of exchange is observed.     

Stereochemical course of thiophosphoryl transfer catalyzed by cytosolic phosphoenolpyruvate carboxykinase

Rat liver cytosolic phosphoenolpyruvate carboxykinase (PEPCK) utilizes inosine 5'-(3-thiotriphosphate) (ITP gamma S) as an excellent substrate, with Km and V values of 0.08 mM and 37 mumol min-1 (mg of protein)-1, respectively, compared with the corresponding values of 0.168 mM and 76 mumol min-1 (mg of protein)-1 for ITP. Thus, the V/Km values for the two substrates are the same. Reaction of (RP)-[gamma-18O2]ITP gamma S with oxalacetate catalyzed by cytosolic PEPCK produces (SP)-thio[18O]phosphoenolpyruvate. Therefore, thiophosphoryl transfer catalyzed by this enzyme proceeds with overall inversion of configuration at P. The reaction mechanism involves an uneven number of phosphotransfer steps, most likely a single step transfer between bound substrates. The results do not support the involvement of a phosphoryl enzyme intermediate in the mechanism.     

Mechanistic studies of phosphoenolpyruvate carboxylase from Zea mays utilizing formate as an alternate substrate for bicarbonate.

Formate is an alternate substrate for bicarbonate in the reaction with PEP catalyzed by phosphoenolpyruvate carboxylase from Zea mays, producing formyl phosphate and pyruvate. The Km for formate is 25 +/- 2 mM, and the maximum velocity is 1% of that for bicarbonate at pH 8.0. Use of [18O]formate produces inorganic phosphate containing 1 equiv of 18O, but no label is incorporated into residual phosphoenolpyruvate. PEP carboxylase catalyzes the hydrolysis of phosphoglycolate or L-phospholactate 2000 times more slowly and D-phospholactate 4000 times more slowly than the reaction between bicarbonate and PEP.     

Isotope trapping and positional isotope exchange with rat and chicken liver phosphoenolpyruvate carboxykinases

Isotope-trapping studies of the enzyme.MgGTP complex were carried out with rat liver cytosolic and chicken liver mitochondrial phosphoenolpyruvate carboxykinases. For the rat liver enzyme, MgGTP was partially trapped from both E.MgGTP and E.MgGTP.OAA complexes, consistent with a steady-state random mechanism. For the chicken liver enzyme, MgGTP was 100% trapped from the E.MgGTP.OAA complex, consistent with a steady-state ordered mechanism. The rate constants for the interaction of MgGTP with the free enzymes are approximately 10(7) M-1 S-1, somewhat lower than the diffusion limit for association. The dissociation rate for the enzyme.MgGTP complexes is 26-92 s-1, reflecting a tightly bound complex with high commitment to catalysis in the presence of oxaloacetate. Positional isotope-exchange studies were also carried out with phosphoenolpyruvate carboxykinases from rat and chicken. No exchange if the beta gamma-18O in [beta gamma-18O, gamma-18O3]GTP to form [beta-18O, gamma-18O3]GTP was detected in the absence of oxaloacetate. In the presence of oxaloacetate, no positional isotope exchange of [beta gamma-18O, gamma-18O3]GTP was detected during initial rate conditions. The results indicate that at least one of the products dissociates rapidly from the E.MgGDP.PEP.CO2 complex relative to the net rate of MgGTP formation from the E.MgGDP.PEP.CO2 complex. A rapid equilibrium between the central complexes in which the beta-phosphoryl of GDP is restricted with respect to torsional rotation cannot be excluded but is unlikely on the basis of the relative rates of catalysis and torsional rotation. The addition of Mn2+, an activator of phosphoenolpyruvate carboxykinase, did not influence the positional isotope-exchange results.(ABSTRACT TRUNCATED AT 250 WORDS)     

Subunit interaction during catalysis. ATP modulation of catalytic steps in the succinyl-CoA synthetase reaction

A new approach for assessing of catalytic cooperativity may occur between subunits has been applied to succinyl-CoA synthetase. This is based on the extent of oxygen exchange between medium [18O]Pi and succinate per molecule of ATP cleaved during steady state succinyl-CoA synthesis. Suitable traps are used to remove succinyl-CoA and ADP as soon as they are released to the medium. With the Escherichia coli enzyme, which has an alpha 2 beta 2 structure, a pronounced increase in oxygen exchange per ATP cleaved occurs as ATP concentration is lowered. In contrast, when the CoA concentration is varied, the oxygen exchange per molecule of product formed remains constant. Also, with the pig heart enzyme, which is shown to retain its alpha beta structure during catalysis and thus has only one catalytic site, no modulation of oxygen exchange by ATP concentration is observed. These experimental findings show that the binding of an ATP either promotes the dissociation of bound succinyl-CoA or decreases its participation in exchange. Measurement of the distribution of [18O]Pi species found as exchange occurs shows that only one catalytic sequence is involved in exchange at various ATP concentrations. These observations along with other controls and results eliminate most other explanations of the ATP modulation of the exchange and suggest that binding of ATP at one catalytic site promotes catalytic site promotes catalytic events at an alternate catalytic site.     

Occurrence and significance of oxygen exchange reactions catalyzed by mitochondrial adenosine triphosphatase preparations.

The capacity of various ATPase preparations from beef heart mitochondria to catalyze exchange of phosphate oxygens with water has been evaluated. Oligomycin-sensitive ATPase preparations retain a capacity for considerable intermediate Pi equilibrium HOH exchange per Pi formed during ATP hydrolysis at relatively high ATP concentration (5 mM). Submitochondrial particles prepared by an ammonia-Sephadex procedure with 5 mM ATP showed more rapid ATPase, less oligomycin sensitivity, and less capacity for intermediate exchange. With these particles, intermediate Pi equilibrium HOH exchange per Pi formed was increased as ATP concentration was decreased. The purified, soluble ATPase from mitochondria catalyzed little or no intermediate Pi equilibrium HOH exchange at 5 mM ATP but showed pronounced increase in capacity for such exchange as ATP concentration was lowered. The ATPase also showed a weak catalysis of an ADP-stimulated medium Pi equilibrium HOH exchange. The results support the alternating catalytic site model for ATP synthesis or cleavage. They also demonstrate that a transmembrane protonmotive force is not necessary for oxygen exchange reactions. At lower ATP concentrations, ADP and Pi formed at a catalytic site appear to remain bound and continue to allow exchange of Pi oxygens until ATP binds at another site on the enzyme.     

Kinetics and compartmentation of energy metabolism in intact skeletal muscle determined from 18O labeling of metabolite phosphoryls

Analyses of isolated intact diaphragm muscle show that at rest only about 30% of the total cellular Pi is metabolically reactive as indicated by 18O incorporation from [18O]water, whereas up to 90% becomes metabolically active incrementally with contractile frequency. Kinetics of [gamma-18O]ATP appearance show that about 90% of the cellular ATP is metabolically active and suggest slowly and rapidly metabolizing compartments of ATP in resting muscle and only rapidly metabolizing compartments in contracting muscle. Rates of [18O]creatine phosphate [( 18O]CrP) appearance are consistent with creatine kinase-catalyzed phosphoryl exchange functioning in an obligatory phosphoryl shuttle system. In noncontracting muscle, ATP turnover rate was 83 nmol.mg protein-1.min-1, and the P/O ratio was determined to be 3.2. ATP utilization increases in direct proportion to contractile frequency with each contracture consuming the equivalent of 0.96 nmol of ATP.mg protein-1 or 2.5-3.5 molecules of ATP/myosin active site. Basal concentrations of nucleotide polyphosphates are not altered when ATP utilization rates increase during contraction. At high contractile frequencies, decreases in CrP concentration occur, but this accounts for less than 4% of total high energy phosphoryls consumed. If metabolic intermediates are free in the aqueous cellular cytosol, each twitch contracture would result in a decrease in ATP concentration of no more than 2% and increases in ADP and AMP concentrations of less than 20 and 7%, respectively. Thus, changes in metabolite concentration must be highly localized or metabolic regulation can be accomplished by a nonallosteric mechanism.     

Dihydrofolate synthetase and folylpolyglutamate synthetase: direct evidence for intervention of acyl phosphate intermediates

The transfer of 17O and/or 18O from (COOH-17O or -18O) enriched substrates to inorganic phosphate (Pi) has been demonstrated for two enzyme-catalyzed reactions involved in folate biosynthesis and glutamylation. COOH-18O-labeled folate, methotrexate, and dihydropteroate, in addition to [17O]-glutamate, were synthesized and used as substrates for folylpolyglutamate synthetase (FPGS) isolated from Escherichia coli, hog liver, and rat liver and for dihydrofolate synthetase (DHFS) isolated from E. coli. Pi was purified from the reaction mixtures and converted to trimethyl phosphate (TMP), which was then analyzed for 17O and 18O enrichment by nuclear magnetic resonance (NMR) spectroscopy and/or mass spectroscopy. In the reactions catalyzed by the E. coli enzymes, both NMR and quantitative mass spectral analyses established that transfer of the oxygen isotope from the substrate 18O-enriched carboxyl group to Pi occurred, thereby providing strong evidence for an acyl phosphate intermediate in both the FPGS- and DHFS-catalyzed reactions. Similar oxygen-transfer experiments were carried out by use of two mammalian enzymes. The small amounts of Pi obtained from reactions catalyzed by these less abundant FPGS proteins precluded the use of NMR techniques. However, mass spectral analysis of the TMP derived from the mammalian FPGS-catalyzed reactions showed clearly that 18O transfer had occurred.     

Relationship of the intra- and extramitochondrial adenine nucleotide ratios during synthesis of phosphoenolpyruvate using extramitochondrial ATP

Mitochondria prepared from the livers of guinea pig, chicken, and pigeon all actively synthesize phosphoenolpyruvate from oxalacetate and GTP, utilizing phosphoenolpyruvate carboxykinase. It was previously shown (Wilson, D. F., Erecińska, M., and Schramm, V. L. (1983). J. Biol. Chem. 258, 10464-10473) that phosphoenolpyruvate carboxykinase is freely reversible and that, in conjunction with nucleoside diphosphate kinase and malate dehydrogenase, which are also present in the mitochondria, it can be used to measure the intramitochondrial [ATPfree]/[ADPfree]. In this study, synthesis of phosphoenolpyruvate by guinea pig liver mitochondria was studied under conditions for which the only source of GTP was extramitochondrial ATP via adenine nucleotide translocase and nucleoside diphosphate kinase (the mitochondria were treated with rotenone, oligomycin, uncoupler, and fluorocitrate). When the extramitochondrial [ATP]/[ADP] was greater than the intramitochondrial [ATPfree]/[ADPfree] calculated from the phosphoenolpyruvate carboxykinase reaction, there was net synthesis of phosphoenolpyruvate, but when it was less, there was net disappearance of phosphoenolpyruvate. Thus, the intramitochondrial [ATPfree]/[ADPfree] was equal to the extramitochondrial value at the point of reversal of the phosphoenolpyruvate carboxykinase reaction. This equality of the intra- and extramitochondrial adenine nucleotide ratios occurred with a measured mitochondrial membrane potential of approximately -36 mV, whereas in the previous experiments, equality was observed for conditions in which the measured membrane potential was -111 to -125 mV. Thus, adenine nucleotide translocation was not dependent on the transmembrane electrical potential and must, therefore, have occurred by electroneutral exchange.