Control of Phosphoenolpyruvate Synthesis by Substrate Level Phosphorylation in Guinea Pig Liver Mitochondria*

SUMMARY Control of P-enolpyruvate synthesis has been studied in guinea pig liver mitochondria with respect to the relative rates of GTP generation directly from substrate level phosphorylation or indirectly from ATP via nucleoside diphosphate kinase. With malate or pyruvate plus malate as substrates, P-enolpyruvate production was greatly decreased in State 3 by addition of fluorocitrate (an inhibitor of aconitase) to prevent flux through substrate level phosphorylation. On the other hand, when ATP production by the electron transport chain was prevented by use of an uncoupling agent (together with oligomycin to inhibit ATPase activity), rates of P-enolpyruvate production from cY-ketoglutarate were the same as in State 3, and were also unaBected by addition of fluorocitrate. In the absence of oxidative phosphorylation and with fluorocitrate present together with malate or pyruvate plus malate, addition of exogenous ATP was much less effective than oc-ketoglutarate in promoting Penolpyruvate synthesis. Intramitochondrial GTP levels and rates


SUMMARY
Control of P-enolpyruvate synthesis has been studied in guinea pig liver mitochondria with respect to the relative rates of GTP generation directly from substrate level phosphorylation or indirectly from ATP via nucleoside diphosphate kinase.
With malate or pyruvate plus malate as substrates, P-enolpyruvate production was greatly decreased in State 3 by addition of fluorocitrate (an inhibitor of aconitase) to prevent flux through substrate level phosphorylation.
On the other hand, when ATP production by the electron transport chain was prevented by use of an uncoupling agent (together with oligomycin to inhibit ATPase activity), rates of P-enolpyruvate production from cY-ketoglutarate were the same as in State 3, and were also unaBected by addition of fluorocitrate.
In the absence of oxidative phosphorylation and with fluorocitrate present together with malate or pyruvate plus malate, addition of exogenous ATP was much less effective than oc-ketoglutarate in promoting Penolpyruvate synthesis.
Intramitochondrial GTP levels and rates of P-enolpyruvate formation were consistently lower in State 4, State 3, and the uncoupled plus oligomycin state with pyruvate plus malate than with a-ketoglutarate plus malate as substrate when fluorocitrate was present, indicating a limitation in the rate of phosphorylation of GDP by intramitochondrial ATP for optimal rates of P-enolpyruvate synthesis.
The maximal rate of conversion of ATP to GTP by nucleoside diphosphate kinase in the intact mitochondria was estimated to be 2 to 3 nmoles per min per mg of protein compared with a maximal rate of P-enolpyruvate formation of about 22 nmoles per min per mg of protein obtained with a-ketoglutarate plus malate as substrate. It is concluded that provision of GTP by substrate level phosphorylation is essential to support rates of P-enolpyruvate formation by guinea pig liver mitochondria greater than 3 nmoles per min per mg of protein.
In animal species such as the guiuca pig, rabbit, and human, mitochondria may contain the major part of the total 1'.enolpyruvate carbosykinase (EC 4.1.1.32) activity of the liver. The exact proportion depends on the dietary status since only the cytosolic enzyme is adaptive (l-3).
Earlier studies have shown that with a-ketoglutarate as substrate, P-enolpyruvate production was stimulated both by addition of uncoupling agents and by addition of a phosphate acceptor system, and that exogenously added ATP was relatively ineffective (4-9).
Ishihara and Kikuchi (9) noted that under conditions of maximal P-enolpyruvate synthesis in guinea pig liver mitochondria, an approximately equal amount of cY-ketoglutarate was oxidized, and suggested that substrate level phosphorylation was preferred over ATP as an energy donor due to the low activity of nucleoside diphosphate kinase.
However, Garber and 13allard (10) challenged this view on the basis of their data, also with guinea pig liver mitochondria, which showed that addition of 2,4-dinitrophenol to uncouple oxidative phosphorylation decreased the rate of P-enolpyruvate formation relative to the rate obtained with State 3 respiration when a variety of precursors were used. Furthermore, addition of cr-ketoglutarate did not relieve the inhibition of I'-enolpyruvate formation induced by the uncoupler.
An apparent equilibration of the ATP and GTP pools was also observed, such that the GTP levels appeared to correlate well with both AT?? formation and P-enolpyruvate synthesis.
The main reason for the discrepancies in results obtained by the different workers may relate to the fact that uncoupling agents not only release respiratory control in the electron transport chain but also progressively stimulate a latent ATPase in isolated mitochondria before inhibitory effects are observed with increasing uncoupler concentrations (1 l-14). Thus, with the relatively high uncoupler concentration used bj Garber and Ballard (lo), utilization of GTP for P-enolpyruvate synthesis would compete with its hydrolysis by the combined actions of nucleotide diphosphate kinase and the activated ATPase.
In the present experiments, use of an uncoupling agent to prevent ATP formation by the electron transport chain and induce a high respiratory activity and an oxidized state of the intramitochondrial pyridine nucleotide has been combined with the simultaneous addition of oligomycin.
This causes an inhibition of respiratory chain-linked ATPase activity (14), and hence eliminates any possible drain on GTP via ATP formation through nucleoside diphosphate kinase. In addition, metabolism of substrates through specific portions of the citric acid cycle has been defined by use of fluorocitrate to inhibit aconitase (15) and 2thenoyltrifluoroacetone to inhibit succinate dehydrogenase (16). Thus, in the presence of fluorocitrate and with malate or pyruvate plus malate as substrates, effects of externally added ATP can be investigated without the complication of GTP formation at the substrate level phosphorylation step. Likewise, when succinate dehydrogenase is inhibited by 2-thenoyltrifluoroacetone, cu-ketoglutarate can be used to generate GTP without providing a source of intramitochondrial malate. Measurements of flux through P-enolpyruvate carboxykinase have been combined with direct assays for intramitochondrial ATP, GTP, CoA, succinyl-CoA, malate, and P-enolpyruvate.
The present data clearly support the conclusions of Ishihara and Kikuchi (9) and others (4-8) that direct formation of GTP by succinate thiokinase is indeed essential to support maximal rates of P-cnolpyruvate formation by isolated guinea pig liver mitochondria.
They also illustrate the alternative control of P-enolpyruvate formation by intramitochondrial oxalacetate availability, which will be described more thoroughly in a subsequent publication.

EXPERIMESTAL PROCEDURES
Mitochondria were prepared from livers of male albino guinea pigs (250 to 300 g in weight) by a minor modification of the method of Schneider and Hogeboom (17). The isolation medium contained 225 mM mannitol, 75 mM sucrose, and 0.1 mM EDTA.
Titrations of carbonyl cyanide p-trifluoromethoxyphenylhydrazone concentration against the respiratory rate were performed with mitochondria incubated at the same protein concentration as used in the experiments in order to ascertain the correct concelltration for maximal stimulation of respiration and complete loss of ATP formation by oxidative phosphorylation.
This was determined separately from measurements of glucose-6-P formation by the glucose-hrsokinase trapping system. All reactions were started by addition of substrate solution to the reaction rnedium preincubated for 1 min with mitochontlria (3 to 5 mg of protein per ml) in a chamber maintained at 30". Oxygen was blown over the surface of the stirred reaction medium.
For measurements of the total content of metabolites in the reaction medium, l-or 2.ml aliquots were rernovrtl at 2-min intervals up to 10 rnin, and the protein was precipitated with 140; (w/v) perchloric acid. The supernatant after ccntrifugation was neutralized with 6 N K&03.
In some experiments, mitochondria were separated from the incubatioil medium by rapid centrifugation through silicone oil as described previously (18).
Metabolites were determined by ffuorornetric enzyme procedures as described by Williamson and ('orkcy (19). A'I'P and GTP were assayed in the same cuvette by taking advantage of the different specificities of hexokinase and P-glyceratc kinase for ATP and GTP (19) specificities under conditions of the assays showed that at COIF centrations of 5 PM, ITP and ATP but not GTP or UTP reacted with hexokinase, while ATP and GTP but not UTP reacted with P-glycerate kinase. Thus, BTP in the sample was first reacted by following the increase of NADPH fluorescence in a reaction system containing NADP, glucose, hexokinase, and glucose-6-P dehydrogenase.
Subsequently, NADH and 3-P-glycerate were added and GTP was assayed by following the decrease of NADH fluorescence upon addition of glyceraldehyde-3-P dehydrogenase and P-glycerate kinase. Succinyl-CoA was assayed by following the fluorescence decrease of NADH upon addition of succingl-CoA: acetoacetatetransferase (EC 2.8.3.5) in a reaction medium containing acetoacetate and fl-hydroxybutyryl-CoA dehydrogenase (EC 1. 1.1.35) (20). a-Ketoglutarate was assayed fluorometrically in the presence of excess aspartate and NADH by coupling aspartate aminotransferase with malate dehydrogenase in an analogous manner to the method described for the aspartate assay (19). P-Enolpyruvate was assayed by coupling pyruvate kinase with lactate dehydrogenase (10). Protein was determined by the biuret method as described by Clelantl alIt Slater (21).

RESULTS
AND DISCUSSION Table I shows a comparison of rates of P-enolpyruvate synthesis and respiration by guinea pig liver mitochondria incubated with different substrates in the presence and absence of fluorocitrate under conditions of State 3 and uncoupled plus oligomycin respiration.
These data illustrate the restriction imposed on the rate of P-enolpyruvate formation when GTP can only be generated from ATP, and in contrast to the results of Garber and Ballard (10) reveal the importance of direct GTP formation by substrate level phosphorylation for optimal rates of P-enolpyruvate synthesis from malate.
With malate as sole substrate, Y-enolpyruvate formation was inhibited about 70% relative to the State 3 rate of 6.7 nmoles per min per mg of protein by addtion of either uncoupler plus oligornycin or fluorocitrate. Similar effects were observed with pyruvatc plus malate as substrates ( Table I). In the presence of uncoupler plus oligomycin, ATP formation by oxidative phosphorylation is prevented, while fluorocitrate addition stops the generation of GTP by substrate level phosphorylation in the citric acid cycle. When the mitochondrial ATPase was allowed to remain active by addition of uncoupler alone, the rate of P-enolpyruvate synthesis from malate fell to 0.6 nmoles per min per mg of protein compared with the rate of 2.3 nmoles per min per mg of protein observed in the uncoupled plus oligomycin state. The occurrence of flux into the citric acid cycle and subsequent substrate level phosphorylation with malate as sole substrate may be inferred from the high ratios of oxygen consumption to I'-enolpyruvate formation shown in Table I, and the accumulation of citrate in the medium upon fluorocitrate addition.
Presumably acetyl-CoA required for citrate formation is provided by the oxidation of endogenous fatty acids (22). Since addition of fluorocitrate in State 3 diminished P-enolpyruvate formation from 6.7 to 2.2 nmoles per min per mg of protein without affecting the intramitochondrial ATP levels, the latter value would appear to represent the maximal rate of GTP formation from ATP via nucleoside diphosphate kinase. The decreased rates of P-enolpyruvate formation observed in the uncoupled and uncoupled plus oligomycin states relative to State 3 may be accounted for partly by the absence of GTP formation from ATP and partly by a decreased flux through substrate level phosphorylation induced by ATP depletion (cf. Fig. 1) which diminishes aotivation of endogenous fatty acids to the CoA esters (22). With fluorocitrate present in addition to uncoupler and oligomycin, with malate or pyruvate plus malate as substrates, the generation of both ATP and GTP is prevented, and the rate of P-enolpyruvate synthesis falls essentially to zero. In contrast to the above data, when a-ketoglutarate was used as substrate, neither fluorocitrate nor uncoupler plus oligomycin produced a significant inhibition of P-enolpyruvate formation (Table I). However, although not shown in Table I Inhibition of the ATPase by oligomycin increased the intramitochondrial ATP levels to 4 nmoles per mg of protein (cj. Fig. 2) compared with values of 0.9 nmoles per mg of protein obtained with uncoupler alone. That the increased ATP level is the result of an altered balance between its rate of production from GTP and its hydrolysis by ATPase, rather than due to residual oxidative phosphorylation, may be deduced from the very low rates of glucose-6-P formation obtained when the glucose-hexokinase trapping system was included in the uncoupled and uncoupled plus oligomycin states. This interpretation is also supported by the observation that with pyruvate plus malate as substrate in the presence of fluorocitrate, addition of oligomycin to the uncoupled state failed to increase intramitochondrial ATP from the low level of 0.5 nmoles per mg of protein.
When malate was included with o(ketoglutarate as substrate, rates of P-enolpyruvate formation were increased 3-to 4-fold (Table I) and were linear over a 30-min interval.
Rates obtained in the uncoupled plus oligomycin state were the same as in State 3, and no statistically significant effect of fluorocitrate was observed in either state. Furthermore, rates of P-enolpyruvate formation with uncoupler alone were only slightly lower than the State 3 rate, probably because the high flux through substrate level phosphorylation diminished the importance of the energy drain through the ATPase.
Measurements of oc-ketoglutarate uptake in the uncoupled plus oligomytin state with fluorocitrate present showed that malate addition increased the uptake from 8 to 18 nmoles per min per mg of protein. Under the latter conditions, the net reaction appears to be that of a stoichiometric conversion of a-ketoglutarate to P-enolpyruvate with utilization of 3 oxygen atoms per mole of P-enolpyruvate produced (cj. Table I).
The present data are in close agreement with those of lshihara and Kikuchi (9) and confirm their conclusion that the production of GTP via substrate level phosphorylation is essential to support high rates of I'-enolpyruvate synthesis by isolated guinea pig liver mitochondria.
Furthermore, when similar experimental conditions are compared, there is no disagreement with the data reported by Garber and Hallard (10). However, disagreement is apparent in the interpretation of results obtained by use of uncoupling agents, particularly with malate as substrate. This arises partly because of the energy drain on GTP (via nucleoside diphosphate kinase) caused by high mitochondrial ATPase activity at uncoupler concentrations beyond t.he minimum necessary to inhibit osidative phosphorylation, and partly to the presence of unrecognized substrate level phosphorylation. Con sequently, in this paper little reliance is placed on interpretation of data obtained by the use of uncoupler alone. When the two variables of ATl'ase activity and substrate level phosphorylation derived from intramitochondrially generated a-ketoglutarate are controlled by use of oligomycin and fluorocitrate, the data become more readily interpretable and clearly reveal a limitation in the rate of production of GTP from ATP by nucleosidc diphosphate kinase.
In another set of experiments, effects of externally added ATP on P-enolpyruvate synthesis were compared with those of a-ketoglutarate addition (Table II). The incubation medium contained fluorocitrate, uncoupler, and oligomycin in order to prevent generation of intramitochondrial ATP by oxidative phosphorylation, and also to prevent substrate level phosphorylation except when cy-ketoglutarate was present as an exogenous substrate.
The data show that although flux through P-enolpyruvate carboxykinase was stimulated appreciably by exogenously added ATP with malate or pyruvate plus malate as substrates, the stimulation by a-ketoglutarate was 5-to 6-fold greater. The effect of ol-ketoglutarate addition on P-enolpyruvate formation was smaller with pyruvate plus malate as substrate than with malate alone, which probably reflects competition between P-enolpyruvate carboxykinase and citrate synthase for intramitochondrial oxalacetate.
Likewise with pyruvate alone as substrate, the rate of P-enolpyruvate formation induced by Qketoglutarate addition was small as a result of the low rate of oxalacetate generation by pyruvate carboxylase.
With 2-thenoyltrifluoroacetone also present to inhibit the formation of intramitochondrial oxalacet,ate from the added a-ketoglutarate, the stimulatory effects of both AT1 and a-ketoglutarate addition were diminished.
However, ol-ketoglutarate remained far more effective than ATP in promoting P-enolpyruvate synthesis, again suggesting that the rate of production of GTP from ATP via nucleoside diphosphate kinase in guinea pig liver mitochondria is unable to support a rate of P-enolpyruvate formation greater than 2 to 3 nmoles per min per mg of protein, which is 10 to 15% of the observed maximal rate.
In order to substantiate the correlation between substrate level phosphorylation and flux through P-enolpyruvate carboxykinase, ATP and GTP levels were determined in mitochondria incubated with either pyruvate plus malate or a-ketogluatrate plus malate as substrates under conditions of State 4, State 3, and uncoupled plus oligomycin respiration (Fig. 1). With fluorocitrate included in the reaction medium, generation of GTP in the presence of pyruvate plus malate could only occur from intramitochondrial ATP via nucleoside diphosphate kinase. In State 4, when rates of P-enolpyruvate production were low because of oxalacetate limitation resulting from the high NADH :NAD ratio characteristic of this state, ATP levels remained high (5 to 7 nmoles per mg of protein) with both substrates, but P-enolpyruvate production and GTP levels were considerably lower in the absence of flux through the substrate level phosphorylation step. However, in State 4 with pyruvate plus malate as substrate, limited oxalacetate as well as limited GTP availability may contribute toward restricting the rate of P-enolpyruvate formation to the observed low rate of 1 nmole per min per mg of protein.
This is because NADH levels were higher in the presence of pyruvate plus malate (1.26 nmoles per mg of protein) than with ar-ketoglutarate plus malate (0.78 nmoles per mg of protein).* The different effectiveness of pyruvate plus malate, and oc-ketoglutarate plus malate as precursors for P-enolpyruvate synthesis was even more marked under conditions of State 3 and uncoupled plus oligomycin respiration.
The levels of NADH remained relatively low and constant (0.3 to 0.5 nmoles per mg of protein), suggesting that differences of oxalacetate availability did not provide an uncon trolled variable.
In the uncoupled plus oligomycin state, GTP levels were hardly detectable with pyruvate plus malate as substrate, and P-enolpyruvate production was negligible. With State 3 respiration, the steady state level of intramitochondrial ATP was lower than in State 4, but despite a high rate of ATP synthesis, the rate of P-enolpyruvate formation was 10 times lower with pyruvate plus malate as substrate t,han with a-ketoglutarate plus malate.
Thus, with no direct generation of GTP by substrate level phosphorylation, the rate of P-enolpyruvate formation by guinea pig liver mitochondria appears to be limited to 1 to 3 nmoles per mg of protein per min by the low activity of nucleoside diphosphate kinase. Within this range, the higher rates appear to correlate with higher intramitochondrial ATP: ADP ratios since rates of I'-enolpyruvate formation with pyruvate plus malate as substrate were higher in State 3 than in the uncoupled plus oligomycin state (Fig. 1). Fig. 2 shows the kinetic changes of metabolites in guinea pig liver mitochondria preincuba.ted with fluorocitrate, carbonyl cyanide P-trifluoromethoxyphenylhydrazone, and oligomycin upon addition of ol-ketoglutarate.
The rates of production of both P-enolpyruvate and malate increased with time while the intramitochondrial GTP level rapidly stabilized to a value of about 0.2 nmoles per mg of protein.
On the other hand, the intramitochondrial levels of ATP and succinyl-CoA decreased appreciably after an initial rise over the first 6 min of incubation. These data illustrate the lack of equilibration between GTP and ATP when there is an energy drain on the system via P-enolpyruvate formation.
The levels of acetyl-CoA remained negligible, and the sum of CoA and succinyl-CoA was constant within experimental error, so that the fall of succinyl-CoA between 6 and 10 min shown in Fig. 2 represents a decrease of the succinyl-CoA : CoA ratio from 0.53 to 0.37. The accumulation of malate and succinate in the medium on addition of P-enolpyruvate indicates that flux through P-enolpyruvate carboxykinase was less than through cY-ketoglutarate dehydrogenase, as also shown directly from measurements of oc-ketoglutarate uptake.
Since the calculated rate of production of GTP was greater than its rate of utilization for P-enolpyruvate synthesis, a flux of 2 to 3 nmoles per min per mg of protein in the direction of GTP to ATP via nucleoside diphosphate kinase must have occurred, thereby accounting for the rise of intramitochondrial ATP. This flux is about equal to the maximal flux through nucleoside diphosphate kinase previously estimated for the reverse direction from ATP to GTP (cf. Table I).
Upon addition of oc-ketoglutarate to the mitochondria, the intramitochondrial levels of malate and P-enolpyruvate fell rapidly, presumably as a result of ol-ketoglutarate entry, but subsequently increased again during the course of the incubation period (Table  III).
The stimulation of I'-enolpyruvate formation with time, therefore, can be accounted for by an increase of the intramitochondrial concentration of oxalacetate. This follows from the fact that the NADH levels remained approximately constant between 0.1 and 0.2 nmoles per mg of protein so that an increased intramitochondrial malate concentration would be expected to yield a proportional increase of oxalacetate concentration. Although this could not be detected by direct assay, an estimate can be made assuming equilibrium of malate dehydrogenase, a constant pH, and an NAD :NADH ratio of 200 characteristic of a highly oxidized state of the mitochondrial pyridine nucleotide system (23). For an intramitochondrial malate concentration of 340 PM (Table III), the estimated oxalacetate concentration is 2 PM, which is below the lowest reported value of 9 PM for the K, of mitochondrial P-enolpyruvate carboxykinase for oxalace- tate (24). The observed rate of P-enolpyruvate formation was 8 nmoles per min per mg of protein, which is about one-third the maximal rate observed with a-ketoglutarate as substrate at saturating extramitochondrial malate concentration, in accordance with the postulate that when GTP formation is not rate-limiting, P-enolpyruvate formation is a function of the oxalacetate concentration.
The sequence of events resulting from the increased availability of oxalacetate to P-enolpyruvate carboxykinase may be interpreted as follows.
Increased demand for GTP by P-enolpyruvate synthesis presumably causes a fall of the guanine nucleotide phosphate potential, thereby resulting in the observed fall of succinyl-CoA levels (Fig. 2), which was reproducible in a number of experiments.
The fall of succinyl-CoA was accentuated at higher malate concentrations, since at 1 rnhf extramitochondrial malate, succinyl-CoA levels of 0.2 nmoles per mg of protein were obtained under the same incubation conditions with cr-ketoglutarate as substrate.
Since succinyl-CoA is an inhibitor of isolated oc-ketoglutarate dehydrogenase (25,26),3 the fall of the succinyl-CoA : CoA ratio produces an increased flux through this step in the intact mitochondria, so that an increased rate of GTP production is able to keep pace with the new rate of P-enolpyruvate formation.
Deinhibition of cY-ketoglutarate dehydrogenase probably results in a fall of the intramitochondrial a-ketoglutarate concentration, thus accounting for the increased rate of ol-ketoglutarate influx observed with a rise of the extramitochondrial malate concentration.
Intramitochondrial oc-ketoglutarate concentrations could not be measured directly in these experiments because of the high concentration in the medium, but assuming a similar concentration gradient across the mitochondrial membrane as that of malate because of their similar charge (27), a fall of intramitochondrial oc-ketoglutarate concentration with time is predicted from the data in Table III.
The observed fall of the succinyl-CoA levels without an apparent change of GTP content (Fig. 2) provides an unexpected result. Calculations of the GTP phosphate potential by assumillg equilibrium at succinate thiokinase give values of 6.9 X lo4 liters mole-' for rat heart mitochondria incubated with acetylcarnitine plus malate in State 4 (26), while values an order of magnitude lower were calculated for rat liver mitochondria under conditions of uncoupled plus oligomycin respiration with a-keto glutarate as substrate.3 Differences between the two sets of values were caused primarily by a higher succinate content in the liver mitochondria.
With an inorganic phosphate concentration of about 20 mM in the mitochondrial matrix, minimal values for the GTP:GDP ratio of about 100 are estimated. This implies that the intramitochondrial GDP concentration is very low and that even without changes of the phosphate concentration, large changes of the GTP phosphate potential could be produced without this being detected from analytical measurements of the GTP content.
It would appear, therefore, that the succinyl-CoA: CoA ratio is a much more sensitive indicator of the GTP:GDP ratio than the GTP level. On the other hand, analytically determined ATP:ADP ratios provide values of about 5 or below, which makes the ATP level more responsive to the mitochondrial energy state than the GTP level. Since the equilibrium constant for nucleoside diphosphate kinase is about unity (28), it is clear that large differences between ATP : ADP and GTP : GDP ratios in the mitochondrial matrix are incompatible with the concept of rapid equilibration between ATP and GTP pools (10). Results presented in this paper demonstrate the close interaction between the GTP requirements for mitochondrial P-enolpyruvate synthesis and its rate of production by substrate level phosphorylation.
Flux through nucleoside diphosphate kinase from ATP to GTP in guinea pig liver mitochondria appears inadequate to support rates of P-enolpyruvate formation greater than 2 to 3 nmoles per min per mg of protein, so that the possibility of rates of P-enolpyruvate formation significantly greater than flux through a-ketoglutarate dehydrogenase is precluded. A considerably lower rate of P-enolpyruvate formation than flux through ar-ketoglutarate dehydrogenase is possible, however, because of independent control factors affecting the availability of intramitochondrial oxalacetate. This is made possible in terms of GTP metabolism due to the conversion of GTP to ATP by nucleoside diphosphate kinase being kinetically favored over the reverse direction because of the different relative levels of the XTP and XDP nucleotides and the inhibitory effects of MgADP (29,30)