Kinetics and regulation of hepatoma mitochondrial NAD(P) malic enzyme.

Kinetic studies of Morris 7777 hepatoma mitochondrial NAD(P) malic enzyme were consistent with an ordered mechanism where NAD adds to the enzyme before malate and dissociation of NADH from the enzyme is rate-limiting. In addition to its active site, malate apparently also associates with a lower affinity with an activator site. The activator fumarate competes with malate at the activator site and facilitates dissociation of NADH from the enzyme. The ratio of NAD(P) malic enzyme to malate dehydrogenase activity in the hepatoma mitochondrial extract was found to be too low, even in the presence of known inhibitors of malate dehydrogenase, to account for the known ability of NAD(P) malic enzyme to intercept exogenous malate from malate dehydrogenase in intact tumor mitochondria (Moreadith, R.W., and Lehninger, A.L. (1984) J. Biol. Chem. 259, 6215-6221). However, NAD(P) malic enzyme may be able to intercept exogenous malate because according to the present results, it can associate with the pyruvate dehydrogenase complex, which could localize NAD(P) malic enzyme in the vicinity of the inner mitochondrial membrane. The activity levels of some key metabolic enzymes were found to be different in Morris 7777 mitochondria than in liver or mitochondria of other rapidly dividing tumors. These results are discussed in terms of differences among tumors in their ability to utilize malate, glutamate, and citrate as respiratory fuels.

of liver, regenerating liver, and many other organs, but is present in tumor mitochondria in levels proportional to the rapidity of cell division (7). However, in tumor mitochondria, the level of NAD(P) malic enzyme activity is still considerably lower than that of malate dehydrogenase ((S)-malate:NAD+ oxidoreductase, EC 1.1.1.37) (1, [7][8][9]. In spite of this, when malate is supplied to tumor mitochondria, it can be channeled exclusively into NAD(P) malic enzyme so that pyruvate and COz are the major products and both are produced at equivalent rates (1,7).
It is not known if NAD(P) malic enzyme is the same in all rapidly dividing tumors. Therefore, we have made a detailed kinetic study of this enzyme, which we recently purified from the Morris 7777 hepatoma (13), and compared the results with those found previously in less detailed kinetic studies of NAD(P) malic enzyme purified from the 22 aH hepatoma (8).
Also, some important kinetic properties of NAD(P) malic enzyme are compared with those of malate dehydrogenase to further characterize competition between these two enzymes for malate. It can be calculated from previous results that in Ehrlich mitochondria, the specific activity of alanine aminotransferase with pyruvate as a substrate would be at least 5-fold higher than that of the fully activatedpyruvate dehydrogenase complex (1,(14)(15)(16)(17)(18). However, when Ehrlich tumor mitochondria are supplied with glutamate plus malate, even in the absence of ADP, about half of the pyruvate generated by NAD(P) malic enzyme reacts with each of these two enzymes (1). This suggests that the pyruvate dehydrogenase complex may associate with NAD(P) malic enzyme with the result being that pyruvate generated by NAD(P) malic enzyme is a better substrate than free pyruvate for the pyruvate dehydrogenase complex. Therefore, in this paper, we have also investigated interactions between the pyruvate dehydrogenase complex and NAD(P) malic enzyme.

MATERIALS AND METHODS
Enzymes and Reagents-Bovine heart pyruvate dehydrogenase and a-ketoglutarate dehydrogenase complexes, bovine and rat liver mitochondrial glutamate dehydrogenases, aspartate aminotransferase, and Morris 7777 hepatoma mitochondrial NAD(P) malic enzyme were prepared as described previously (13,(19)(20)(21)(22)(23). Pig heart mitochondrial malate dehydrogenase, citrate synthase, and fumarase were obtained from Boehringer Mannheim. Other enzymes, coenzymes, substrates, and reagents were obtained from Sigma. Stock solutions of all reagents used in assays were adjusted to the pH of the assay. Solutions of succinyl-CoA were prepared fresh daily, and the concentrations of succinyl-CoA and acetyl-coA were measured as described previously (19). Methods used to dialyze or chromatograph enzymes on Sephadex G-200 or G-25 and to prepare enzyme for use in these experiments were described previously (19, 24).
Concentrations of Enzyme and Protein-The concentrations of pure pyruvate dehydrogenase complex, a-ketoglutarate dehydrogenase complex, and NAD(P) malic enzyme were measured as described previously with bovine serum albumin as a standard (25, 26). The concentrations of pure glutamate dehydrogenase, malate dehydrogenase, and citrate synthase were measured spectrophotometrically at 280 nm as described previously (19,24). The concentration of protein in mitochondrial extracts was measured by the biuret reaction on trichloroacetate precipitates of the extracts (26).
Obtaining and Analyzing Kinetic Data-Michaelis constants were obtained from the Michaelis-Menten relationship as described previously (19). The effects of allosteric modifiers of NAD(P) malic enzyme in the presence of constant levels of substrates were evaluated with Equation 1 for an activator (A) or Equation 2 in the case of an inhibitor (I) as described previously (27,28).
In Equations 1 and 2, Au is the change in velocity produced by the modifier, V is the velocity in the absence of modifier, VA and Vi are the velocities in the presence of saturating modifier, and KO and K, are the apparent dissociation constants of the modifier. Equations 1 and 2 are empirically identical to the Michaelis-Menten equation. Therefore, data corresponding to Equations 1 and 2 were directly fitted to these equations using the nonlinear regression computer program. Product inhibition data were analyzed using the least squares method as described previously (29). Non-Michaelis-Menten kinetic results were evaluated as described previously (30) and in the text. Each experiment was the average of three experiments. Standard errors were 4 0 % of the mean.
The quite low K, of oxalacetate in the malate dehydrogenase reaction was measured spectrophotometrically at 340 nm with a Cary high-performance spectrophotometer from Varian Associates, Inc.
Preparation and Assay of Enzymes in Mitochondrial Extracts-Mitochondria were isolated from the viable tissue of Morris 7777 hepatomas (initially provided by Dr. Michael Lee, New Jersey Medical School), mitochondrial extracts were prepared, and enzymes were assayed with the same methods (Methods I and I1 (31) plus additional methods) used in the accompanying article to measure enzyme levels in liver mitochondria (31). With the exceptions of the enzymes noted below, all extraction procedures yielded comparable levels of enzyme activity.
Because relatively large amounts of protein (-0.5 mg) were required for assay of alanine aminotransferase activity in the tumor extracts, it was centrifuged briefly at 3000 x g to reduce its turbidity. Alanine aminotransferase activity was not detected in the 3000 X g precipitate.
In addition to Method I. assays for NAD:isocitrate dehydrogenase were also performed with mitocGondria that were homogenizedin 0.1 M potassium phosphate (pH 7.6), 1 mM ADP, 1 mM dithiothreitol, 0.1 mM EDTA, 10 pg/ml leupeptin, and 0.5% Triton X-100. The activity of NAD:isocitrate dehydrogenase was the same in liver mitochondria as reported previously (31), but the activity in hepatoma mitochondria was -1.2-fold lower.
Assays for NAD(P) malic enzyme were also performed with mitochondria that were homogenized in 50 mM Bes/triethanolamine (pH 7.2), 1 mM dithiothreitol, 1 mM EDTA, 0.5% Triton X-100, 10 pg/ml leupeptin, and 5 mM fumarate. Similar results were obtained with mitochondria prepared with this and the previously described method (Method I) (31).
For assays of NAD(P) malic enzyme, the medium contained the same buffer used for preparing the mitochondrial extract together with 5 PM rotenone, 1.0 mM NADP, 10 mM malate, and 5 mM MgCl,. The reaction was started with Mn2+ after a constant rate ( 4 0 % that with MnZ+) was established in its absence. Added fumarate had no effect. As a test for the strictly NADP-dependent malic enzyme, the extract was assayed at low concentrations of malate (2 mM) and ' The abbreviations used are: Hepes, N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid; Bes, N,N-bis(2-hydroxyethyl)-2-aminoethanesulfonic acid. NADP (0.2 mM) in the presence of 1 mM ATP and 2 mM MgCl,, conditions under which NAD(P) malic enzyme is largely inactive. In agreement with previous studies (7) of Morris 7777 hepatoma mitochondria, there was no NADP malic enzyme activity.
Other enzyme assays were performed as described previously (31), and specific activities are expressed as micromoles of product/minute/milligram of protein. The forward malate dehydrogenase reaction was assayed in mitochondrial extracts in the presence of 1.0 mM NAD, 1.0 mM malate, 0.1 mM acetyl-coA, and 20 rg/ml citrate synthase as described previously (24). With the exception of the alanine aminotransferase reaction (which, as mentioned above, was assayed in a centrifuged extract), all enzyme assays and measurements of protein concentrations were performed with uncentrifuged extracts.

Regulation of NAD(P) Malic Enzyme by Substrates and
Fumarate-A series of kinetic experiments were performed under conditions (50 mM Tris-HC1, 10 mM MnC12, 0.1 mM EDTA, and 0.5 mM dithiothreitol (pH 7.4)) comparable to those employed previously in studies of mitochondrial NAD(P) malic enzyme purified from 22 aH hepatoma mitochondria (8). Under these conditions and when NAD was the coenzyme, double-reciprocal plots of the velocity of Morris 7777 mitochondrial NAD(P) malic enzyme uersus malate concentration were not linear unless the allosteric activator fumarate was also present (Fig. 1). These results were consistent with Equation 3, the rate equation for Mechanism I, where malate (B) is bound to the active site of the oxidized coenzyme-enzyme complex (E.A); and in the absence of fumarate, malate is also bound with a considerably lower affinity to an allosteric activator site. Fumarate (F), however, competes with and thereby prevents binding of malate at the activator site so that, in the presence of fumarate, doublereciprocal plots of velocity uersus malate concentration are linear.  In Equation 3, u is the initial velocity; [Eo] is the milligrams of enzyme/milliliter, and dissociation constants and rate constants for product-yielding steps are represented by K and k, respectively, as shown in Mechanism I. The solid curve ( Fig. 1, curue A ) , obtained in the presence of NAD and absence of fumarate, was calculated from Equation 3 with [F] = 0, K1/K2 being negligible, and the other constants equal to the values shown in Table I, where Vl, Vz, and V3 equal the maximal specific activities of the steps indicated by kl, kp, and k3, respectively, of Mechanism I. When fumarate is saturating (curue B ) , Equation 3 can be simplified to Equation 4, an equation of the Michaelis-Menten form.
The values of the constants of Equation 4 are also shown in Table I. Thus, according to these results (Fig. l), binding of either malate or fumarate to the activator site increases velocity at the active site, but the effect of fumarate is 1.8fold greater ( V2 = 10, V3 = 18) (Table I). Fumarate produced a comparable increase (1.5-fold) in the K, of malate at the active site (K1 = 0.8 mM, K7 = 1.2 mM) ( Table I).
In the presence of high levels of malate, double-reciprocal plots of velocity uersus NAD concentration were linear in either the presence or absence of fumarate (Fig. 2). Fumarate again increased the maximal velocity -1.5-fold ( Vz = 11, V3 = 16) ( Fig. 2 and Table I), but fumarate had no marked effect on the K,,, of NAD ( Ka = 0.028 to 0.024 mM) ( Fig. 2 and Table  I).
When NADP was the coenzyme, double-reciprocal plots of velocity versus malate concentration were linear in both the presence and absence of fumarate (Fig. 3). Fumarate markedly decreased the K,,, of malate, but had little effect on the V,,, ( Fig. 3 and Table I). These results indicate that in the absence of fumarate, NADP does not readily support binding of malate to the enzyme. Consequently, the K , of malate at the active  site is high (Ki = 12 mM) (Table I), and binding of malate to the activator site was not detected. Fumarate, however, now mainly activated by enhancing binding of malate to the active site (K7 = 1.2 mM). Fumarate had no significant effect on the maximal velocity ( V3 = 15, Vl = 17) (Table I).
Although NADP was not very reactive in the absence of fumarate, in the presence of fumarate, NADP was essentially as active as NAD. In both cases, the V,,, was 15-18 and the K,,, of malate was 1.2-1.4 mM (Table I). However, in the presence of both coenzymes and fumarate, the velocity and effect of lactate dehydrogenase on the reaction were about the same as in the presence of NAD alone (Table II), indicating that as with other NAD(P) malic enzymes, NAD is the preferred coenzyme (8, 33).

TABLE I1
Activity of tumor mitochondrial NAD(P) m a l i c enzyme with NAD versus NADP Assays were performed in the presence of 10 mM malate, 5 mM fumarate, a 1.0 mM concentration of each coenzyme, and, where indicated, 5 pg/ml lactate dehydrogenase (LDH). Other conditions were the same as described in the legend to Table I Reaction in Potassium Phosphate-The same type of kinetic pattern as described above was also found when tumor malic enzyme was assayed in potassium phosphate plus 2 mM MgClz at pH 7.0 (Table I11 and data not shown). Again, fumarate increased the linearity of double-reciprocal plots of velocity uersus malate concentration when NAD was the coenzyme, but now fumarate also slightly decreased the K,,, of malate at the active site (decreased K J (Table 111). Plots of the reciprocal of velocity uersus NAD concentration were also linear in phosphate buffer; and again, fumarate did not markedly alter the K,,, of NAD, but mainly activated by increasing the maximal velocity (Table 111). The reaction with NADP was also consistent with Michaelis-Menten kinetics in phosphate buffer; and again, the main effect of fumarate was to markedly decrease the K, of malate. The major difference between the NAD(P) malic enzyme under the two different assay conditions was found when NAD was the coenzyme and fumarate was absent. Under these conditions, the maximal velocity, when malate was bound only to its active site (VI = 0.16 uersus 5), and the affinity of malate for its allosteric site (K3 = 100 uersus 8) were both considerably lower in phosphate plus M e (Table 111) than in Tris plus Mn2+ (Table I).
Regulation by Product-As shown in Figs. 4-7, in the presence of fumarate, Mn2+, and Tris buffer, NADH was competitive with NAD and noncompetitive with malate. Plots of the reciprocal of velocity uersus NADH concentration intersected at low levels of NADH, and the intersection points are shown in Table IV Table IV). Also, in the absence of fumarate, plots of velocity uersus NADH concentration deviated from linearity (NADH became more inhibitory) as the level of NAD was decreased (Fig. 8).
In phosphate buffer plus Mf, NADH was also quite inhibitory, and NADPH was equally inhibitory. Again, inhibition was reduced by fumarate (Table V). Even 1.0 mM levels of pyruvate were not very inhibitory (Table V).
Regulation by Activators or Inhibitors-Succinate was also an activator, but was less effective than fumarate (Ka higher and maximal activation lower) ( Fig. 10 and Table VI). Oxalacetate was an inhibitor (Ki = 0.18 mM), and saturating oxalacetate inhibited 100% (Fig. 11).
The rate equation for a mechanism where two modifiers compete for the same site can be obtained from Mechanism I and Equation 3 by substituting the second modifier (in this case, succinate or oxalacetate) for malate at the allosteric site (steps indicated by K2 and K3). According to the rate equation for this mechanism, fumarate should increase the apparent dissociation constant of the second modifier, but should have    18 2 5 0 0 0 0 no effect on velocity when the second modifier is saturating and displaces fumarate from the enzyme. In agreement with this mechanism (Fig. 10, curve A ) 10. Plots of specific activity of NAD(P) malic enzyme versus modifier concentration. Fumarate and succinate were the varied modifiers used in curues A and C, respectively. Curve A shows the results obtained when the concentration of succinate was increased in the presence of 5.0 mM fumarate. Remaining experimental conditions were 1.0 mM NAD, 1.0 mM malate, 20 mM potassium phosphate, 2 mM MgC12, 0.1 mM EDTA, and 0.5 mM dithiothreitol (pH 7.0) at 25 "C. The solid curues were calculated with the use of Equation 2 and the values of the constants shown in Table VI.

TABLE VI
Modifier constants of tumor mitochondrial NAD(P) malic enzyme The constant K refers to the activator or inhibition constant. The terms V, V, and V, , ' are the velocity in the absence of modifier, velocity in the presence of saturating modifier, and velocity in the presence of 5.0 mM fumarate plus saturating second modifier, respectively. The values of the constants were obtained from the results shown in Figs. 10 Table VI. to the same site of the same enzyme-substrate or enzymeproduct complex; and consequently, both fumarate and oxalacetate can simultaneously associate with the enzyme.
All other metabolites tested were either inhibitory or had no effect (Table V). At 0.2 mM levels, only succinyl-CoA was inhibitory in the presence or absence of fumarate, and acetyl-CoA was inhibitory in the presence of fumarate. At 1.0 mM levels, ATP and, to a lesser extent, ADP were inhibitors in the presence or absence of fumarate.
Malate Dehydrogenase Versus Malic Enzyme Activity-As shown in Table VI1 and previously (7,9,13), in Morris 7777 mitochondria and in mitochondria of other rapidly dividing tumors, malate dehydrogenase activity is in vast excess over that of NAD(P) malic enzyme. Under conditions more physiological than those used to obtain the results shown in Table  VII, the ratio of malate dehydrogenase to NAD(P) malic enzyme in Morris 7777 mitochondria would be extremely high. Using the results shown in Table VI1 plus the assays of pure NAD(P) malic enzyme described in this paper, it can be calculated that in the presence of 1.0 mM NAD, 1.0 mM malate, 5.0 mM fumarate, 2.0 mM MgC12, 0.1 mM acetyl-coA, and 20 pg/ml citrate synthase (pH 7.0) at 25 "C, the specific activity of NAD(P) malic enzyme in the tumor mitochondrial extract would be 0.04 nmol/min/mg of mitochondrial protein.
Since fumarate and MgC12 have little effect on malate dehydrogenase (24), the specific activity of malate dehydrogenase in the extract under these conditions (citrate synthase plus acetyl-coA present to react with oxalacetate) would be essentially equal to that shown in Table VI1 (Method 111) or 4 orders of magnitude higher than that of NAD(P) malic enzyme. Although the activity of NAD(P) malic enzyme would be -%fold higher if Mn2+ were substituted for M$+ (7) and would also be higher at a higher pH or temperature, increasing the pH and temperature also increases malate dehydrogenase activity (Table VII). However, pyruvate is not a potent inhibitor of NAD(P) malic enzyme (Table V), whereas the Ki of oxalacetate from NADH:malate dehydrogenase at a physiological p H is so low that it is difficult to measure directly with conventional methods (24). The Ki of oxalacetate can, however, be estimated by taking advantage of the fact the malate dehydrogenase reaction is an Ordered Bi Bi reaction, where NAD adds before malate, and the rate-limiting step is dissociation of NADH from the enzyme (34,35). According to the rate equation for this mechanism (36), when NADH is saturating and malate is added as a product inhibitor, the reciprocal of the Ki of oxalacetate can be obtained from the intersection point of double-reciprocal plots of velocity versus oxalacetate concentration. These plots, in the presence of 400 PM NADH (which was found independently to be a level of NADH -10-fold higher than its K,) and 0-6.0 mM malate, are shown in Fig. 12. The intersection point corresponds to a Ki of oxalacetate of 5 PM. Inhibition by malate in these assays was specific in that even 10 mM levels of aspartate had no effect on malate dehydrogenase activity (data not shown).
In addition to being inhibited considerably less by its keto acid product, NADH (especially in the presence of fumarate), citrate, and glutamate were considerably less potent inhibitors of NAD(P) malic enzyme ( Table V) than was found to be the case previously with malate dehydrogenase (20, 24, 33, 35). However, physiological levels of these inhibitors in addition to oxalacetate would not be expected to reduce malate dehydrogenase activity in tumor mitochondria to a level even close to that of NAD(P) malic enzyme. Furthermore, the fact that malate dehydrogenase is considerably more reactive than NAD(P) malic enzyme with malate generated internally by fumarase (1) in tumor mitochondria, even in the absence of low energy or State 3 conditions, indicates that malate dehydrogenase is not inhibited to the extent that it is less active than NAD(P) malic enzyme.
Although the kinetic properties of malate dehydrogenase described above were obtained with malate dehydrogenase purified from normal mitochondria, there is apparently no Leuek of enzymes in liuer and Morris 7777 hepatoma mitochondrial extracts Mitochondrial extracts were prepared and assays were performed as described under " Materials and Methods" and in Ref. 31. For some enzymes, there was a wider range of values in the hepatomas than in liver. Therefore, the ranges of values in the hepatomas are shown. 'The assay was not actually performed, but was calculated from the ratio of the performed assays to the indicated assay with pure enzyme. PDHC -ME 0 major kinetic difference between rat liver and hepatoma malate dehydrogenases. In both cases, we found the K,,, of malate, the ratio of activity of the forward and reverse reaction, and inhibition by citrate and glutamate to be the same (data not shown). Effect of Pyruvate Dehydrogenase Complex-As shown in Table VIII, the pyruvate dehydrogenase complex activates NAD(P) malic enzyme. In these assays, acetyl-coA was not present, malate was not oxidized when malic enzyme was not added, and the level of the pyruvate dehydrogenase complex required for activation (20 pg/ml or 2.8 nM) was quite low (Table VIII). Therefore, the increase in malic enzyme activity apparently results from binding of the large pyruvate dehydrogenase complex (M, = 7 x lo6) (37) to NAD(P) malic enzyme ( M , = 2.4 x lo5) (8,13) and is not a consequence of the pyruvate dehydrogenase complex reacting with NAD plus pyruvate, a contaminating malate-oxidizing enzyme, or the pyruvate dehydrogenase complex decreasing inhibition of the malic enzyme by its products. Activation was specific in that glutamate dehydrogenase, the a-ketoglutarate dehydrogenase complex, and citrate synthase were not activators ( Table  VIII).
Activation of NAD(P) malic enzyme by the pyruvate dehydrogenase complex could also take place in Morris 7777 mitochondria. Although citrate synthase also associates with the pyruvate dehydrogenase complex (38), it did not prevent the pyruvate dehydrogenase complex from activating NAD(P) malic enzyme. Furthermore, it can be calculated from the specific activity of the pure enzymes and the data of Table VI1 (using previously described methods) (31) that the level of NAD(P) malic enzyme would be 1.5 mg/ml or 5.6 ~L M and that the level of the pyruvate dehydrogenase complex would be 3 mg/ml or 0.4 PM in the tumor mitochondria. These concentrations are considerably higher than those used in our experiments; and assuming multiple malic enzyme-binding sites on the extremely large pyruvate dehydrogenase complex, the level of this complex could be sufficiently high to associate with a significant fraction of the NAD(P) malic enzyme.
Levels of Other Enzymes in Morris 7777 Mitochndria- The mitochondrial levels of enzymes that could play a role in the oxidation of glutamate and malate in Morris 7777 hepatomas are shown in Table VII. It can be seen that the level of the pyruvate dehydrogenase complex was about the same in both the hepatoma and rat liver, but the activities of enzymes that catalyze the reactions of the Krebs cycle from fumarate to succinyl-CoA, especially isocitrate dehydrogenases, were significantly higher in the hepatoma (aconitase, which is difficult to assay in extracts, was not measured). Alternatively, pyruvate carboxylase was almost absent; and the levels of activities of enzymes involved in glutamate-NH: metabolism, such as aspartate aminotransferase, glutamate dehydrogenase, ornithine transcarbamylase, and especially alanine aminotransferase, were significantly lower in Morris 7777 mitochondria than in liver mitochondria (Table VII). It was demonstrated previously (39) that the level of carbamyl-phosphate synthase I is also quite low in hepatoma mitochondria. The differences found between the levels of enzyme activities in hepatoma versus liver mitochondria (Table VII) were not artifacts produced by the procedures employed to prepare and extract mitochondria. If this were the case, then it would not be expected that the ratio of enzyme activity in liver mitochondria to that in hepatoma mitochondria would be high in some cases and low in others. Furthermore, comparable ratios were found when enzymes were extracted and assayed with several different methods (see "Materials and Methods" and Table 11). The major method used (Method I) to prepare mitochondria does not rupture the liver outer mitochondrial membrane, and electron micrographs of hepatoma mitochondria prepared with Method I (data not shown) also revealed that digitonin removed the bulk of the contaminating vesicles without rupturing the hepatoma outer mitochondrial membranes.
Based upon the citrate synthase assay, we recovered 58 mg of mitochondrial protein/g of hepatoma homogenate protein with a 54% recovery. This gives a value of 107 mg of mitochondrial protein/g of homogenate protein, which would be lower than the value other investigators found with liver (40) and half the value we found with liver (31), but consistent with the observation that hepatomas have %fold less mitochondria than liver (41).

DISCUSSION
Previous kinetic studies of cytosolic malic enzymes were consistent with an ordered mechanism being the predominant reaction pathway for the forward reaction, as shown in Mech- anism I1 (42,43). The tumor mitochondrial malic enzyme also probably catalyzes a predominantly ordered reaction because it has an extremely low affinity for a malate affinity column unless oxidized pyridine nucleotide is also present to promote binding (13). Furthermore, in the absence of fumarate, the e MECHANISM 11. A, oxidized pyridine nucleotide; B, malate; P, CO,; Q, pyruvate; R, reduced pyridine nucleotide; kl-klo, rate constants for the indicated steps.
K, of malate is considerably lower in the presence of NAD than in the presence of NADP, which indicates that the coenzyme plays a major role in determining the affinity of malate for the enzyme. In addition, in the presence of fumarate, NADH was competitive with NAD and non-competitive with malate, which is consistent with the ordered mechanism (42, 43). The rate-limiting step in the ordered malic enzyme reaction (43), and in many other dehydrogenase reactions, is dissociation of the reduced coenzyme from the enzyme ( 4 ) .
The low Ki of NADH and potent inhibition of the tumor malic enzyme by both NADPH and NADH are consistent with this also being the case with the tumor malic enzyme.
When NAD was the coenzyme, the extrapolated maximal velocity was essentially the same in the presence of infinite NAD and constant high malate or infinite malate and in the presence of constant high NAD, indicating in both cases that the two substrates were saturating. Fumarate had little effect on the K,,, of NAD, but both the K,,, of malate and Vma. were increased -1.5-fold by fumarate. According to the rate equa- activating by enhancing dissociation of NADH (increasing ks so V,,, and KB are both increased 1.5-fold) and increasing kl, the rate of association of NAD with the enzyme, so that KA or 4 / k l is not altered.
In addition to the steps shown in Mechanism 11, malate apparently also associates with an allosteric site, where it could activate by increasing dissociation of reduced coenzyme. Similar kinetic patterns would have been obtained, when the concentration of malate was varied, if there were cooperative interaction between the binding sites on the four enzyme subunits or if, instead of associating with an allosteric site, malate associated with the active site of E . NADH (ER) and enhanced dissociation of NADH. However, the latter mechanism would require that malate can readily associate with the free enzyme ( E ) , which, as mentioned above, is apparently not the case.
NADH is apparently also bound to an inhibitory allosteric site on the enzyme because when the level of NAD was low, there was considerably more inhibition by NADH than could be accounted for on the basis of a single NADH-binding site. However, binding of either fumarate or malate to their allosteric site apparently inhibits binding of NADH to its allosteric site. Consequently, in the presence of fumarate, plots of the reciprocal of velocity versus NADH concentration in the presence of various levels of NAD were linear. In addition, when fumarate was absent and the level of NAD was sufficiently high to prevent binding of NADH to its active site (but not to its allosteric site), plots of the reciprocal velocity versus NADH concentration were linear in the presence of various concentrations of malate, and malate was almost competitive with NADH. If, however, in the absence of fumarate the level of NAD was decreased, then plots of the reciprocal of velocity versus NADH concentration were not linear, indicating that lowering the level of NAD permitted binding of NADH to the active site and also decreased binding of malate. Decreased binding of malate, in turn, permitted NADH to associate with the NADH allosteric site.
Succinate is also an activator, but is competitive with and less effective than fumarate. All of the other metabolites tested were inhibitors or had no effect. Of these, oxalacetate was the most potent. However, the K i of oxalacetate (140-180 p~) is probably higher than its free mitochondrial level (44), especially in the Morris 7777 hepatoma, which lacks pyruvate carboxylase (Table VII) and channels exogenous malate into NAD(P) malic enzyme. Alternatively, the level of succinyl-CoA required for inhibition (0.2 mM) is lower than its level in liver mitochondria (0.3-1.4 mM) (45).
Differences among NAD(P) Malic Enzymes-In previous studies of the adrenal cortex mitochondrial NAD malic enzyme (46, 47) and several tumor (22 aH, Ehrlich, and AS-30D) NAD(P) malic enzymes (8), no evidence was presented that suggested an allosteric malate site. Fumarate appeared to lower the K,,, values of malate and NAD with the adrenal NAD malic enzyme, and the main effect of fumarate with the tumor NAD(P) malic enzyme was to produce a marked (5-10-fold) decrease in the K,,, values of malate, NAD, and NADP (8,46,47). On the basis of these kinetic experiments and the behavior of the 22 aH enzyme on an affinity column, investigators (8) concluded that the main action of fumarate is to enhance binding of coenzyme and malate to the enzyme. These results are considerably different from those of our study of the Morris 7777 mitochondrial NAD(P) malic enzyme under comparable assay conditions. In this case, when NAD was the coenzyme, fumarate did not markedly change the K,,, of NAD and slightly increased the K,,, of malate. However, the K,,, values of malate and NAD (0.8 and 0.028 mM, respectively) with the Morris 7777 enzyme were lower than the K,,, values of malate and NAD (3.6 and 0.055 mM, respectively) with the other tumor NAD(P) malic enzymes (8). Although product inhibition and detailed studies of the effects of fumarate and other regulators have only been performed in our study, these results indicate that when NAD is the coenzyme, the other tumor enzymes have a comparatively lower affinity for NAD and, in turn, malate and that consequently, the main effect of fumarate is to decrease the K, values of NAD and malate. In contrast, the main effect of fumarate with the Morris 7777 enzyme is apparently to enhance dissociation of NADH from the enzyme. When NADP is the coenzyme, the Morris 7777 enzyme reacts, as the other tumor enzymes, in that the K,,, of malate is high; there was no evidence of an allosteric malate site, and fumarate markedly decreased the K,,, of malate.
In previous studies, ATP was found to be a competitive inhibitor of malate with both the tumor ( K i = 80 PM) and adrenal cortex (Ki = 0.3 mM) enzymes (8, 46, 47). However, in our study, considerably higher (-1.0 mM) levels of ATP were required for significant inhibition of the Morris 7777 enzyme in the presence and absence of fumarate.
Regulation by Hetero-enzyme Interaction-According to our results, the preferred reactivity of NAD(P) malic enzyme versus malate dehydrogenase with exogenous malate in tumor mitochondria cannot be readily explained on the basis of the kinetic properties of the two enzymes or the levels of activity of the two enzymes in tumor mitochondria. However, a pos-sible explanation is (as we have proposed previously (24, 31, 48) on the basis of experiments performed by us (48) and others (38, 49-52) with pure enzymes) that the only malate dehydrogenase that would be active with NAD plus malate would be the fraction of the total malate dehydrogenase that is not membrane-bound or free but is localized in the matrix as a consequence of associating with membrane-hetero-enzyme complexes between the a-ketoglutarate dehydrogenase complex and aspartate aminotransferase or the pyruvate dehydrogenase complex and citrate synthase. Association of malate dehydrogenase with these hetero-enzyme complexes would provide malate dehydrogenase with a mechanism for the removal of oxalacetate and also results in an increase in malate dehydrogenase activity (24). At a physiological pH, malate dehydrogenase cannot even be readily assayed in the absence of enzymes that react with its products (24). On the other hand, some NAD(P) malic enzymes apparently have a high affinity for the membrane (53); the equilibrium of the malic enzyme reaction is 3 orders of magnitude more favorable than that of malate dehydrogenase (54, 55); and pyruvate is not a potent product inhibitor of NAD(P) malic enzyme (Table V). Therefore, NAD(P) malic enzyme may be localized in the vicinity of the membrane, where its thermodynamic and kinetic properties would enable it to readily intercept malate from malate dehydrogenase as malate passes through the membrane. Furthermore, according to our results, the pyruvate dehydrogenase complex, which has a high affinity for the inner mitochondrial membrane (56), could, by also associating with NAD(P) malic enzyme, not only increase NAD(P) malic enzyme activity, but also play a role in localizing NAD(P) malic enzyme in the vicinity of the membrane.
Energy Metabolism in Morris 7777 Hepatorna-Morris 7777 hepatomas, as other rapidly dividing tumors, can utilize glutamine as a major metabolic fuel (1-5). According to our results, oxidation of glutamate would have several obstacles in Morris 7777 hepatomas because alanine aminotransferase activity is quite low, pyruvate carboxylase activity is almost absent, and the level of aspartate aminotransferase is also significantly lower in Morris 7777 mitochondria than in liver mitochondria. Therefore, in the presence of malate and mitochondrial levels of glutamate, which can be as low as 6.0 mM, and levels of aspartate and a-ketoglutarate, which under the same conditions can be 9.0 and 3.0 mM, respectively (57), aspartate aminotransferase, which can have a K,,, for glutamate as high as 20 mM (23), may not be capable of competing with citrate synthase for oxalacetate. Furthermore, the quite low levels of alanine aminotransferase and pyruvate carboxylase activities plus the adequate level of pyruvate dehydrogenase activity and possibly the observed interaction between the pyruvate dehydrogenase complex and NAD(P) malic enzyme described in this paper would enhance the probability that pyruvate generated by NAD(P) malic enzyme would be utilized to provide citrate synthase with acetyl-coA. In addition, in the presence of low levels of glutamate and low-energy or State 3 conditions, malate dehydrogenase, the pyruvate dehydrogenase complex, and isocitrate dehydrogenases would be quite active; and this plus the high levels of these enzymes and citrate synthase compared with liver mitochondria would promote the synthesis and oxidation of citrate. Consequently, Morris 7777 mitochondria (4), unlike 3924A hepatoma mitochondria (58), do not significantly export citrate for cholesterol synthesis, but instead oxidize citrate, perhaps as a compensatory energy-generating mechanism. Potent inhibition of NAD(P) malic enzyme (our results) and both isocitrate dehydrogenases (59) by the NADPH generated by the high level of NADP:isocitrate dehydrogenase could be overcome by glu-tamate dehydrogenase, which in mitochondria can preferentially react with NADPH (60, 61). In addition, since aketoglutarate generated by NADP:isocitrate dehydrogenase cannot be readily oxidized by the Krebs cycle, but can react with glutamate dehydrogenase (59, 62), the reductive amination of a-ketoglutarate would provide aspartate aminotransferase with the glutamate needed for it to provide the Krebs cycle with a-ketoglutarate.
The reactivity of glutamate dehydrogenase with NADPH and a-ketoglutarate is high under low energy State 3 conditions and considerably more active in State 3 than in high energy State 4 conditions (60). In addition, the level of NH: would be expected to be high in a hepatoma that has a high level of glutaminase (4) and is deficient in carbamyl-phosphate synthase I and ornithine transcarbamylase activities (39). Since the K , of NH: for glutamate dehydrogenase is high and -20 mM (63)) the high level of NH: would further promote the reductive amination of a-ketoglutarate.