Eur. J. Biochem.

68, 551-562 (1976)

Efficiency of Oxidative Phosphorylation and Energy Dissipation by H Ion Recycling in Rat-Liver Mitochondria Metabolizing Pyruvate
+

J6rg W. STUCK1 Pharmakologisches Institut der Universitiit Bern (Received October 22, 1975/Junc 28, 1976)

A method was developed for the calculation of metabolic fluxes through individual enzymatic reactions of pyruvate metabolism including the citric acid cycle in rat liver mitochondria incubated at metabolic states between state 4 and state 3. This method is based on the measurement of the specific radioactivities of the products formed from [2-'4C]pyruvate. With this procedure the energy balance of mitochondria incubated in the presence of [2-'4C]pyruvate, ATP, bicarbonate and phosphate at different ATP/ADP ratios in the medium was calculated. The ATP/ADP ratios were maintained at a steady state with creatine kinase plus creatine as a phosphoroyl acceptor. The calculations revealed that by adding increasing concentrations of creatine up to 20 mM the energy dissipated by the mitochondria decreased but showed a local maximum at 13 mM creatine. Omission of bicarbonate from the medium led to a shift of this maximum. When energy dissipation was minimal the overall P/O ratio was maximal. The amount of energy dissipated was paralleled by the magnitude of the pH gradient across the inner membrane. From these results it was concluded that the recycling of H + ions which consists of a passive leakage of H f ions into the matrix and an active extrusion of these ions out of this compartment, is an important energy dissipating process. The H + ion recycling is thus one of the processes which give rise to the state 4 respiration in mitochondria.

Previous studies have shown that oxidative phosphorylation has a low efficiency in rat liver mitochondria metabolizing pyruvate near state 4 conditions in the presence of ATP, bicarbonate and phosphate [I, 21. In a computer aided study a P/O ratio of only about 0.7 was determined under these conditions [l]. Further investigations have shown that this low efficiency is due to the occurrence of energy dissipating processes which compete with oxidative phosphorylation for the energy supplied by the respiratory chain. One of these energy-dissipating processes was identified as an extramitochondrial ATPase. Inhibiting the effect of this ATPase on energy dissipation with atractylate resulted in a 26 % inhibition of the total energy dissipation [l]. Another energy-dissipating process was identified as the recycling of calcium ions across the inner mitochondria1 membrane. Inhibition of the contribution of this process to the energy dissipation
~

Enzymes. Pyruvate carboxylase (EC 6.4.1.1) ; pyruvate dehydrogenase (lipoate) (NAD') (EC 1.2.4.1); isocitrate dehydrogenase (EC 1.1.1.41); 2-oxoglutarate dehydrogenase (EC 1.2.4.2); malale dehydrogenase (EC 1.1.1.37); 3-hydroxybutyrate dehydrogenase (EC 1.1.1.30); creatine kinase (EC2.7.3.2); citrate synthase (EC 4.1.3.7); hexokinase (EC 2.7.1.1); adenosinetriphosphatase (EC 3.6.1.3); adenylate kinase (EC 2.7.4.3).

with ruthenium red led to a 17'x inhibition of the total energy dissipation [2]. Several reports in the literature indicate that the recycling of H'ions across the inner membrane which consists of a passive leakage of H'ions into the matrix and an active extrusion of these ions out again into the extramitochondrial space is another energy-dissipating process in mitochondria [3 - 51. Unfortunately there is no inhibitor known to abolish the contribution of this process to the total energy dissipation. Therefore it is not possible to measure directly the influence of the Hf recycling on the efficiency of oxidative phosphorylation in coupled mitochondria. However, it can be expected that the H'ion recycling is dependent on the magnitude of the transmembrane protonmotive force since this is the driving force for the leakage of H+ions through the inner membrane. This study attempts to measure the changes in energy dissipation due to H' ion recycling at a steady state with an improved inethod without perturbation of the system under physiological conditions. The transmembrane pH gradient was manipulated by incubating rat liver mitochondria at different ATP/ADP ratios in the incubation medium. The ATP/ADP ratios in the incubation medium were maintained at a steady state

552

Efficiency of Oxidative Phosphorylation

during the incubation period with creatine kinase plus creatine as a phosphoroyl acceptor. This method of varying the transmembrane pH gradients in intact mitochondria is probably better suited to reflect the actual situation in the intact cell than experiments with ionophores or uncouplers. A previously published procedure allowed the calculation of net metabolic fluxes through individual enzymatic reactions of pyruvate metabolism in mitochondria incubated near state 4 conditions [ l ] and was based on the measured specific radioactivities of the metabolites formed from [2-14C]pyruvate. This method suffered from the fact that it was not applicable to mitochondria incubated near state 3 conditions. The reason for this was that only the condensation of oxaloacetate formed by carboxylation of pyruvate or formed from a first passage of citrate through the citric acid cycle could be accounted for in the calculations ; further cycling of citrate through the cycle had to be neglected. The application of this procedure was confirmed experimentally by the observation that near state 4 conditions only a very small amount of citrate was formed in a second condensation of oxaloacetate with acetyl-CoA. On changing the conditions from state 4 to state 3, however, the recycling of oxaloacetate through the cycle increases and can no longer be neglected. In this paper a method for the calculation of metabolic fluxes in mitochondria is presented which is more general and no longer subject to the restrictions mentioned above. With the help of this method the energy balance and the overall P/O ratios of mitochondria metabolizing pyruvate at different extramitochondrial ATP/ ADP ratios were calculated. The paper gives information on the relation between the energy dissipation and the transmembrane pH gradients under these conditions. Part of these data have already appeared in abstract form [6]. MATERIALS AND METHODS
Incubations and Assays with Mitochondria

6.6 mM potassium phosphate buffer pH 7.4, 6.6 mM triethanolamine buffer pH 7.4, 10 mM pyruvate, 250 pg creatine kinase and 0.25 M mannitol/0.07 M sucrose to make a final volume of 3 ml. In incubations without bicarbonate the potassium bicarbonate was replaced by 10 mM KC1. The protein-free neutralized extracts were assayed for pyruvate, malate, citrate, 2-oxoglutarate, acetoacetate, ADP and AMP as previously described [I]. Creatine phosphate was measured with an enzymatic method [9]. Separation of the organic acids by paper electrophoresis and isotope measurements was done as published elsewhere [l]. 3-Hydroxybutyrate and succinate were assayed by isotope methods as described earlier [l] and fumarate as described below. For direct simultaneous measurements of oxygen uptake and redox states of nicotinamide adenine nucleotides or cytochromes, preincubated mitochondria were transferred into a thermostated cuvette in a Hitachi dual wavelength spectrophotometer (model 156). The oxygen tension in the cuvette was measured with a Clark electrode. The suspension in the cuvette was agitated with a magnetic stirrer. For the determination of the transmembrane pH gradients the mitochondria were separated from the incubation medium with the silicone layer method [lo]. A 1 : 1 mixture of Wacker silicone oil AR 150 and AR 100 was used for the separation layer. The volume of the matrix space was determined by measuring the distribution of t3H2]3H20 and ['4C]sucrose. The pH gradients were calculated from the distribution of 5,5-[14C]dimethyl2,4-oxazolylidinedione between the medium and the sucrose-inaccessible space [ l l ] . A l / l (v/v) mixture of Triton X-lOO/xylene supplemented with 4 g/1 Omnifluor was used as a counting liquid for the aqueous solutions.
Calculation of Metabolic Fluxes through Individual Enzymatic Steps in Pyruvate Metabolism and of the Energy Balance of Incubated Mitochondria

Rat liver mitochondria were isolated according to Johnson and Lardy [7] with the exception that the homogenizing medium contained 0.25 M mannitol plus 0.07 M sucrose [8]. Washed mitochondria were suspended in this medium at a concentration of 1 g original liver weight per ml. Mitochondria were incubated in conical flasks at 37 "C as described elsewhere [I] with the exception that split off radioactive 14C02 was not trapped since the measurement of creatine phosphate required immediate neutralization of the incubations stopped with perchloric acid. The medium contained 4 mM ATP, 10 mM MgS04, 10 mM potassium bicarbonate,

Fig. 1 depicts the specific radioactivities of the products formed from [2-14C]pyruvate in the citric acid cycle as expressed in units of the specific radioactivity of the pyruvate. The basic properties of this reaction scheme as well as the reversibilities of the reactions have been discussed previously [l]. By defining X i as the fraction of the total citrate formed in the ith cycle by condensation of oxaloacetate with acetyl-CoA in the citrate synthase reaction the total amount of citrate formed is:
it],,, = X
I

+ X , . . . + X , = ig xi.

(1)

Inspection of the specific radioactivities of the citrates in Fig.1 yields the following expression for the pyr-

J. W. Stucki

553

by 1st cycle by 2nd cycle by 3rd cycle by n t h cycle itz rut liver mituchondr,-iu. Numbers in parentheses Fig. 1. Speci/i'c ,ndiutic,tivities u/ citric. cycle intermediutes formrd,/rotn [ 2 - ' 4 C ] p ~ , r u v u t e designate the specific radioactivities of the products expressed in units of the specific radioactivity of the [2-'4C]pyruvate precursor. For the discussion of the reversibilities of the reactions see [I]. Pyr = pyruvate, AcCoA = acetyl-CoA, OxAc = oxaloacetate, Cit = citrate, Suc = succinate, Fum = fumarate, Ma1 = malate

uvate units of radioactivity of the total amount of citrate (*Cit): *Cit = 2X1

From the relations Eqns (7) and (8) a can be calculated as the solution of the quadratic equation
a '

+ 3X2 + (2 + 1/2)X3+ .

[Cit],,, - E

*Cit 2

+ *Cit

2 [Cit],,,

= 0.

(9)

or by making use of Eqn (1): "Cit

=

2 [Cit],,,

+ X2 + 1/2X3 + . . . 1 +2 -. x n.
2-

(3)

The coefficients of this quadratic equation contain the measured amount and radioactivity of the citrate formed. Once the recycling ratio CI is known XI is obtained from Eqn (8) as
XI = [Cit],,, ( 3
-a).

By defining the recycling of citrate through the cycle by the recycling ratio a as :
a=
x 2 -

(10)

x 1

x 3_ ..
x 2

. -~ _Xn _ .
xn-1

(4)

X , can be expressed as

xl=x,ai-l.

(5)

This definition of CI is valid under the experimental conditions used because the pool sizes of the intramitochondria1 metabolites are orders of magnitudes lower than the net throughputs through the system. Hence deviations from this relationship which are due to the approach of a steady state can be neglected in the overall balance equations. Insertion of Eqn ( 5 ) in Eqn (3) then yields: *Cit
= 2 [Cit],,,

The other values of X i are then obtained by making use of Eqn (4). The amounts of 2-oxoglutarate and succinate formed in the cycle are rather low when compared with the citrate formed (see Results) and therefore the calculation of the origin of these components can be neglected. The origins of the remaining two major metabolites formed in the cycle, malate and fumarate, are calculated as follows. By defining Yo as the fraction of the total malate formed by reduction of oxaloacetate formed via the pyruvate carboxylase and Yi as the fraction of the total malate formed in the ith cycle we obtain : [Malltat= Y o + YI

Y2

+...+ Yn=

,f Y,.(11) =o

aJ + clXl jgoT.
01

(6)

Inspection of Fig. 1 yields then for the radioactivity of malate expressed in units of pyruvate: *Ma1 = Yo

Since obviously a is bounded within the limits 0 I a I1 the sum in Eqn (6) converges and we obtain *Cit = 2[Cit],,,

+ 2Y1 + (1 + 1/2) Y z + ...

(12)

+1
~~

a X1
-

42

(7)

from which one obtains analogous to the calculations for citrate: (1 - 4 2 ) - (13) Yo = [Mal]t,t - ("Ma1 - [Mal]tot) (1 - a ) Y1 = ([Mall,,, - Yo) (1 - E ) . (14)

Application of the above procedure to Eqn (1) yields [Cit],,, =

x 1
~

1-a'

(8)

554

Efficiency of Oxidative Phosphorylation

Again, we have
u=-=--

Yz Yr

Y3 Y2

_..._ - Y, Yn1

can be in oxidation phosphorylation ( [Crt-P]oxphos) calculated from the balance equation : (15 ) [Crt-P]oxphos = [Crt-P]tot - ADPfound - 2AMPround

The other values of Y, can be calculated by making use of Eqn ( 3 5). If it is assumed that malate and fumarate are almost in equilibrium, the amount of fumarate can be calculated from the measured radioactivity of the fumarate and the specific radioactivity of the malate. The origins of fumarate can be calculated with expressions analogous to Eqns (13 - 15). The amounts of pyruvate units condensed to ketone bodies has been calculated as described previously [l]. Since every molecule of oxaloacetate formed by carboxylation of pyruvate must appear as one or the other of the citric acid cycle products the amount of pyruvate carboxylated is calculated as the sum of these products. The amount of pyruvate oxidized is the sum of the pyruvate transformed into ketone bodies and that entering the citric acid cycle via acetyl-CoA. The latter value is obtained as the sum of i times the sum of the products formed in the ith cycle whereby, as stated above, for 2-oxoglutarate and succinate only the first cycle is considered. The total amount of pyruvate used is calculated as the sum of pyruvate carboxylated and that oxidized. It is important to note that the measured amount of pyruvate utilized is not used in the above calculations. Therefore, this parameter serves as a check to test the validity of the analytical determinations and the calculations. Fluxes through the individual reactions of the citric acid cycle were calculated from the value of the pyruvate entering the citric acid cycle via acetyl-CoA and the products formed in the individual reactions of the cycle [I]. The amounts of NADH and FADH2 formed and utilized in the reactions were obtained from these flux data as described earlier [l]. The amounts of NADH and FADH2 oxidized were calculated as the difference of formed and utilized reducing equivalents. By using P/O ratios of 3 and 2 for NADH and FADH2 respectively the energy formed in the respiratory chain can be expressed in energy equivalents of ATP. The energy utilized was calculated as the sum of ATP used for pyruvate carboxylation [l] and that used for creatine phosphate formation. Creatine phosphate is formed from creatine and ATP produced in oxidative phosphorylation and since 4 mM ATP was present in the incubation medium also from the added ATP. To calculate the energy balance of the mitochondria1 reactions it is therefore necessary to correct the total creatine phosphate formation for that formed by the added ATP. The contribution to the total creatine phosphate formation ([Crt-PI,,,) of that portion formed from ATP produced

('6)

which takes into account the conversion of the adenine nucleotides in the adenylate kinase reaction. The pyruvate carboxylase is exclusively localized within the matrix space of the mitochondria [12]. Since the added ATP does not enter this compartment in energized mitochondria [14] no correction for pyruvate carboxylation supported by added ATP is necessary. The energy utilized by the mitochondria was therefore calculated as the sum of [Crt-P]ouphos plus the pyruvate carboxylated. The dissipated energy was calculated and expressed in energy equivalents as the energy produced in the respiratory chain minus the utilized energy as discussed in previous work [l].
Materials

Male Wistar rats with an average weight of 150250 g from the Institut fur Zuchthygiene of the University of Zurich were used. All radioactive chemicals and Omnifluor were obtained from NEN Chemicals GmbH (Dreieichenhain, Germany). Carbonyl cyanide p-trifluoromethoxyphenylhydrazone was a generous gift from Dr Heytler (Dupont, Wilmington, U.S.A.). The specific radioactivity of [2-14C]pyruvate was determined as previously described [ 13. Enzymes and coenzymes were purchased from Boehringer Mannheim GmbH (Mannheim, Germany). Wacker silicone oil was supplied by Stehelin & Co (Basel, Switzerland). All other reagents were of the highest purity commercially available. Aqueous solutions were prepared with twice distilled water. RESULTS. Table 1 summarizes the results of an experiment in which mitochondria were incubated in the presence of [2-'4C]pyruvate, bicarbonate, ATP and phosphate in the medium. Isocitrate, oxaloacetate, aconitate, aspartate and glutamate were also measured with standard enzymatic methods or by isotope measurements. Because these acids were present only in very low concentrations ( < 0.03 pmo1/3 ml) they were not included in the table and neglected in the calculations in Tables 2 and 3. The ATP/ADP ratios in the medium were lowered by additions of nonlimiting amounts of creatine kinase plus increasing concentrations of creatine as a phosphoroyl acceptor. From the measured amounts and radioactivities of the products formed from [2-14C]pyruvate (Table 1) the origins of the metabolites (Table 2) and the energy balance of the experiments (Table 3) were calculated as described under Methods.

J. W. Stucki

555

Table 1. Measured amounts and radioactivities of products,formed f r o m (2-'4C]pyruvute in mitochondria incubated with creatirie Mitochondria from 0.5 g of liver containing 18 mg protein were incubated during 10 min as described under Methods. The [2-'4C]pyruvate used had a specific radioactivity of 71 009 counts x min-' x pmol-'. Results are quoted for 3 ml solutions. The values given are means of duplicate incubations with mitochondria from the same preparation. The radioactivity of the products is given in parentheses Metabolites found Amount and radioactivity with additions of creatine (mM)
~~ ~~ ~~

~~

~~

None

3.3

6.7

10.0

13.3

16.7

20.0

pmol (countsjniin) __
~

~-

~-

~~

Citrate 2-Oxoglutarate Succinate Fumarate Malate Acetoacetate 3-Hydroxybutyrate Pyruvate used Creatine phosphate ADP AMP

4.09 (629 139) 0.24 (24678) (74230) 5.22 (460034) 0.65 (20048) 19.76
-

3 24 (502 212)

0.23
(15 756) (59 548) 3.91 (355 842)

1.05
(26 792) 17.25 7.13 0.25 0.21

2 78 (441 406) 0 20 (15 456) (48 588) 2 94 (271 442) 1 12 (32718) 14.80 12.75

2 28 (367940) 0 21 (1 1596) (39 224) 3 21 (203632) 1.47 (35 344) 13.73 14.35 0.49 0.27

187 (305060) 0.22 (12704) (29907) 1 54 ( 154 956) 1.58 (38 192) 12.48 14.77 0.57 0.30

1.6 (258 472) 0.13 (14 8 12) (23 321) 1.31 ( 135 700) 1.82 (46552) 11.35 15.40 0.71 0.33

1.43 (230069) 0.19 (1 1 570) (1 8 681) 1.03 ( I07 118) 185 (48 404) 11.14 15.62 0.80 0.40

0.21 0.20

0.33
0.22

The results in Table 2 show that the pyruvate carboxylase was successively inhibited by increasing concentrations of creatine. This confirms earlier findings that the activity of this enzyme is essentially dependent on the ATP/ADP ratio in the medium in the presence of saturating pyruvate concentrations [ 141. The production of the ketone bodies was enhanced by this transition from state 4 towards state 3 conditions as was also shown by others [15,16]. On the other hand, the pyruvate units entering the cycle decreased with increasing creatine concentrations. This indicates that the net flux of acetyl units entering the cycle is mainly dependent on the supply of the cycle with oxaloacetate formed in the pyruvate carboxylase reaction. The net flux through the isocitrate-dehydrogenase reaction reflects the amount of acetyl-units oxidized in the citric acid cycle (Table 3). This quantity remained almost constant up to creatine concentrations of 13 mM and then decreased at higher creatine concentrations. The energy balance of the experiment in Table 3 shows that the energy utilized for carboxylation of pyruvate plus formation of creatine phosphate increased only up to a creatine concentration of 6 mM and then remained almost constant upon further addition of creatine. In contrast to this, the energy dissipation first decreased at 3 mM creatine, then increased to a local maximum at 13 mM creatine and decreased again upon further elevation of the creatine concentration. Since more energy is dissipated than utilized by the mitochondria under the present experimental conditions, this peculiar behaviour of energy dissipation is also visible in total energy production as well as in the oxygen consumed by the mitochon-

dria. The oxygen uptake was either calculated as the sum of NADH plus FADH2 oxidized (Table 3) or measured directly with a Clark electrode (Fig. 2). The recycling ratio (Table 2) paralleled the total energy production. This suggests that the recycling of the acids in the citric acid cycle is regulated by the energy demands of the mitochondria. The results in Table 3 demonstrate further that the ratio of ATP utilized/oxygen consumed which is an apparent overall P/O ratio and which reflects the efficiency of oxidative phosphorylation in this system, is inversely related to the changes in energy dissipation : an increase of the energy dissipation was accompanied by a lowering of this ratio and vice versa. To assess the effect of pyruvate carboxylation which consumes ATP within the matrix space, the mitochondrial energy utilization was also studied in the absence of this reaction in the following experiments. Since pyruvate is not carboxylated in the absence of added bicarbonate these experiments were therefore performed without bicarbonate. In the experiment depicted in Fig. 2 mitochondria were preincubated at different creatine concentrations both in the presence and absence of bicarbonate for 10 min. After this preincubation the rate of oxygen uptake was measured with a Clark electrode. Before the solutions became anaerobic the reactions were stopped by the addition of perchloric acid and the formed phosphocreatine was measured in the deproteinized and neutralized extracts. Since in the absence of bicarbonate in the medium essentially no pyruvate is carboxylated the formation of phosphocreatine is the only energy-utilizing reaction which occurs under these conditions. In contrast, as stated in Methods

556

Efficiency of Oxidative Phosphorylation

Table 2 Measured and calculated amounts of metabollteJ from dlfjerent ongin5 produced during mcubation of rat liver mitochondria wlth creatine Values given were either taken from Table 1 or calculated from the results of Table 1 as indicated under Methods Metabolites Amount with additions of creatine (mM) __ _ _ _ ~ -____ None 33 67 pnol
~

___
16 7 20 0

10 0

13 3

. .~

Citrate total 1st cycle 2nd cycle 3rd cycle 4th cycle 5th cycle 2-Oxoglutarate Succinate Fumarate total Reduction 1st cycle 2nd cycle 3rd cycle Malate total Reduction 1st cycle 2nd cycle 3rd cycle 4th cycle Acetoacetate 3-Hydroxybutyrate Pyruvate carhoxylated Oxidized total Oxidized to ketone bodies Oxidized via cycle Pyruvate utilized calc. Measured Recycling ratio
ct

4.09 3.33 0.62 0.12 0.02

-

3.24 2.57 0.54 0.11 0.02

-

2.78 1.99 0.57 0.16 0.05 0.01 0.20 0.10 0.52 0.33 0.14 0.04 0.01 2.94 1.88 0.76 0.21 0.07 0.02 1.12 0.23 6.54 8.58 2.70 5.88 15.12 14.80 0.285

2.28 1.50 0.51 0.18 0.07 0.02 0.21 0.07 0.40 0.22 0.13 0.04 0.01 2.1 1 1.15 0.63 0.22 0.07 0.04 1.47 0.25 5.07 8.85 3.44 5.41 13.92 13.73 0.343

1.87 1.12 0.45 0.18 0.08 0.04 0.22 0.08 0.30 0.14 0.10 0.04 0.02 1.54 0.69 0.52 0.21 0.08 0.04 1.58 0.27 4.01 8.67 3.70 4.97 12.68 12.48 0.399

1.60 1.04 0.36 0.13 0.05 0.02 0.13 0.09 0.22 0.08 0.09 0.03 0.02 1.31 0.55 0.49 0.18 0.07 0.02 1.82 0.33 3.35 8.33 4.30 4.03 11.68 11.35 0.349

1.43 0.96 0.32 0.10 0.03 0.02 0.19 0.07 0.18 0.08 0.07 0.02 0.01 1.03 0.43 0.40 0.13 0.04 0.02 1.85 0.34 2.90 7.76 4.38 3.38 10.66 11.14 0.332

0.24 0.16 0.85 0.63 0.18 0.03 0.01 5.22 3.82 1.14 0.21 0.04 0.01 0.65 0.14 10.56 8.98 1.58 7.40 19.54 19.76 0.185

0.23 0.10
0.66 0.46 0.16 0.03 0.01

3.91 2.67 0.99 0.20 0.04 0.01 1.05 0.19 8.14 8.67 2.48 6.19 16.81 17.25
~

0.206

both the formation of phosphocreatine and the carboxylation of pyruvate are energy-utilizing reactions when bicarbonate is added to the medium. Moreover, it is important to realize that it is no longer possible to apply the procedure for the calculation of net metabolic fluxes given under Methods when no pyruvate is carboxylated. Therefore in the absence of bicarbonate oxygen uptake cannot be calculated from the isotope data but must be measured directly. The measured oxygen uptake of the mitochondria incubated with bicarbonate was essentially the same as that calculated from the results in Table 3. The rate of oxygen uptake was lower without than with bicarbonate. Furthermore, the qualitative behaviour of the oxygen uptake with and without bicarbonate was different, the most obvious difference being a shift of the maximal oxygen consumption to 6.6mM creatine when bicarbonate was omitted. The amount

of phosphocreatine formed was higher in the presence of bicarbonate than in its absence. Since without bicarbonate in the medium no pyruvate is carboxylated within the matrix space the supply of the citric acid cycle with oxaloacetate is even more limited than in the incubations with bicarbonate and high creatine concentrations. This suggests that under these conditions the production of reducing equivalents in the citric acid cycle may become a limiting factor for the respiratory rate at elevated ADP concentrations in the medium. This assumption was tested in the experiment shown in Fig. 3. In this experiment the redox state of the nicotinamide adenine nucleotides of preincubated mitochondria was estimated by measuring the changes in absorbance increase at 340- 375 nm upon transition from the aerobic to the anaerobic state (Fig. 3 A). With the same procedure the redox state of cytochrome b was

J. W. Stucki

551

Table 3. Energy halance of mitochondria incubated with creatine From the results in Table 2 net fluxes of metabolites through individual enzymatic reactions were calculated as noted under Methods. From these values the total amounts of reducing equivalents produced and utilized were then listed in this table. For the calculation of energy produced, utilized and dissipated as well as for the overall P/O ratio see Methods and text. The same effects were observed in four other experiments Metabolites Amount with addition of creatine (mM) __ None 3.3 6.7 10.0 pmol
-~
_.
~~~~~

~~

..

13.3

16.7

20.0

NADH produced total viu pyruvate deliydrogenase via isocitrate dehydrogenase via oxoglutarate dehydrogenase via malate dehydrogenase NADH utilized total via malate dehydrogenase v i a 3-hydroxybutyrate dehydrogenase FADHz produced NADH oxidized FADHz oxidized Energy produced (in ATP equivalents) Energy utilized (in ATP equivalents) Energy dissipated (in ATP equivalents) Overall P I 0

16 65 8 98 3 31 3 07 1 29 4.59 4.45 0.14 2.91 12.06 2.91 42.00 10.56 31.44 0.71

15 58 8 67 2 95 2 78 118 3.32 3.13 0.10 2.62 12.26 2.62 42.02 15.20 26.82 1.02

16 13 8 58 3 10 2 90 155 2.44 2.21 0.23 2.80 13.69 2.80 46.67 18.52 28.15 1.12

16 61 8 85 3 13 2 92 171 1.62 1.37 0.25 2.85 14.99 2.85 50.67 18.39 32.28 1.03

1644 8 67 3 I0 2 88 179 1.10 0.83 0.27 2.80 15.34 2.80 51.62 17.61 34.01 0.97

14 38 8 33 2 43 2 30 I 32 0.96 0.63 0.33 2.21 13.42 2.21 44.68 17.38 27.30

~~~

12 46 7 76 195 176 0 99 0.85 0.51 0.34 1.69 11.61 1.69 38.21 16.92 21.29
..
~ ~ ~

1.11

I .27

estimated by measuring the changes in absorbance at 430-410 nm after transition of the aerobic to the anaerobic state (Fig. 3 B). In the absence of bicarbonate the mitochondria1 nicotinamide adenine nucleotides were in a more oxidized state than in its presence, notably at creatine concentrations above 6 mM. This indicates that the production rate of reducing equivalents is low when pyruvate is the only substrate, i.e. when the pyruvate carboxylase is not active. A comparison of the oxygen uptake rates in Fig. 2 with the curves in Fig. 3 A shows that the redox state of the nicotinamide adenine nucleotides is practically the same in mitochondria incubated with and without bicarbonate at creatine concentrations below 6 mM. However, the rate of oxygen uptake is much higher in the presence of bicarbonate than in its absence. This discrepancy may be explained by the fact that pyruvate carboxylation acts as an intramitochondrial ATPase and hence activates respiration. Moreover, it emerges from the figures that the occurrence of a maximal respiratory rate at 13 mM creatine in mitochondria incubated with bicarbonate cannot be the cause of an increased supply of the respiratory chain with reducing equivalents since this maximum was not accompanied by a similar change in the redox state of the nicotinamide adenine nucleotides. The redox changes of cytochrome b in Fig. 3 B did not show such pronounced differences as those

of the nicotinamide nucleotides. In the presence of bicarbonate cytochrome b was only slightly more reduced than in its absence even at elevated creatine concentrations. Moreover, the cytochrome h did not elicit a change in the redox state which could explain the maximum in oxygen uptake at 13 mM creatine in the incubations with bicarbonate. From the measurements shown in Fig. 3A and 3 B it can thus be concluded that the behaviour of the oxygen uptake and the energy production as shown in Table 3 and Fig.2 are not regulated by the availability of reducing equivalents in the respiratory chain but rather by the energy demands of the mitochondria. The results shown in Table 3 revealed that an increase in oxygen uptake may not only result from an enhanced rate of energy utilization but also from an increased rate of energy dissipation. Several reports in the literature indicated that the recycling of H 'ions across the inner membrane is an energy-dissipating process [3- 51. This process is due to the leakage of H 'ions into the matrix down their electrochemical gradient and an energy-utilizing pumping of these ions again out of the matrix and it can be assumed that the transmembrane electrochemical gradient of H+ions acts as a driving force for this leakage. The determination of the electrochemical potential difference of H + ions across the inner membrane requires both the measurement of the membrane

558

Efficiency of Oxidative Phosphorylation

15

t 10
c m

>--

6 5

0
[Creatine] ( m M )

Fig. 2. Measured oxygen uptake i0,o) and creatine phosphate (m, 0 ) formation of rat liver mitochondria incubated with (0,m] and without (0, 0)bicarbonate. Mitochondria from 0.3 g of liver containing 10 mg of protein were preincubated during 10 min as described under Methods. After this preincubation the reaction mixture was transferred into thermostated vessels where the oxygen uptake rate was measured with a Clark electrode as described under Methods. This measured rate was then converted into units of patom 0 consumed in 10 min per 18 mg protein for direct comparison with the results in Table 3. The total amount of creatine phosphate measured after stopping the reactions was corrected to give the total amount of creatine phosphate formed expressed as pmol formed in 10 min per 18 mg protein. The points give the mean values of duplicate incubations with mitochondria from the same preparation. The same effects were observed in four other experiments

potential as well as of the pH gradient across the inner membrane. The membrane potential can only be determined in the presence of valinomycin in the incubation medium whereas the pH gradient can be measured by the distribution of the weak acid 5S-dimethyl-2,4-oxazolylidinedione [5,17,18]. In a typical experiment carried out exactly as the control incubation in Table 1 without creatine but with 0.2 FM valinomycin present in the medium, the amount of pyruvate carboxylated was only 0.23 pmol/ 10 min and the intramitochondrial ATP/ADP ratio was 0.12 under these circumstances. Earlier experiments have shown that the intramitochondrial ATP/ ADP ratio in the control incubation is around 0.7 [l,21. Furthermore, the pyruvate carboxylase activity is strongly dependent on this ratio in the presence of saturating pyruvate concentrations [14]. Therefore the 98% inhibition of the pyruvate carboxylase by valinomycin is probably mediated by a fall of the intramitochondrial ATP/ADP ratio. Since the metabolic fluxes and the energy balance of the incubations can be measured with the [2-14C]pyruvate method outlined in the Methods only when a sizable amount of pyruvate is carboxylated, these measurements are no longer possible in the presence of valinomycin. Hence the membrane potential cannot be measured in our experiments with the only currently available method. On the other hand, the measurement of the transniembrane pH gradient by the distribution of 5,5-['4C]dimethyl-2,4-oxazolylidinedione did not influence the activity of the pyruvate

. '
._
0
7 m 3
0 '

loo?

0 '

0 0
10

1 0
[Creatine] ( m M )

20

20

[Creatine] ( m M )

Fig. 3. Steady .state oxidation of nicotinamide adenine nucleotides and of cytochrome b in incubated mitochondria. Mitochondria from 0.2 g of liver containing 7 nig protein were preincubated during 10 min as described under Methods. After this preincubation the redox state of the nicotinamide adenine nucleotides (A) and of cytochrome h (B) was estimated as described in the text. The measured values were expressed as a percentage of the absorbance changes measured in a transition from the aerobic to the anaerobic state of mitochondria incubated with 1 pM carbonylcyanide p-trifluoromethoxyphenylhydl-azone without added pyruvate. The points give the mean values of duplicate incubations with mitochondria from the same preparation. The same effects were observed in six other experiments

J. W. Stucki

559
0.6
A

carboxylase as was investigated in earlier experiments (J. W. Stucki, unpublished results). It is important to realize that the trdnsmembrane pH gradients reflect only the chemical but not the electrochemical potential difference of H +ions across the inner membrane. Therefoi-e the measurement of the pH gradient gives no exact value of the total driving force of the H + leakage process [19]. But the magnitudes of the p H gradients can still be used as an approximation of this driving force under the conditions used in our experiments (see Discussion). The values of the pH gradients of mitochondria incubated with and without bicarbonate at different creatine concentrations in the medium were calculated from the distribution of the weak acid 5,5-dimethyl2,4-oxazolylidinedione and are depicted in Fig. 4A. A comparison of the amount of energy dissipated shown in Table 3 with the magnitude of the transmembrane pH gradients in Fig. 4 A showed that these two quantities are closely related: an increase in the pH gradient is accompanied by an increase in energy dissipation and vice versa. This result strongly suggests that the H + ion recycling across the inner membrane is an important energy-dissipating process under the experimental conditions used here. Since net metabolic fluxes cannot be measured with the [2-'4C]pyruvate method in incubations without bicarbonate as discussed above no direct comparison of energy dissipation with the p H gradients is possible under these conditions. But since, as also shown in earlier studies [I 1, in the presence of 4 mM ATP in the medium more energy is dissipated than utilized by the mitochondria the energy dissipation is a dominant process for the regulation of the respiratory rate. Hence also the parallelism of the respiratory rates in Fig. 2 and the pH gradients in Fig. 4 A strongly suggests that H i ion recycling is an important energydissipating process both in the presence and absence of bicarbonate. The energy dissipating H ion recycling should also influence the overall efficiency of oxidative phosphorykdtion. Therefore changes of the pH gradient across the inner membrane should change the overall P/O ratios of oxidative phosphorylation. The overall P/O ratios in Fig.4B were taken from the results in Table 3 for incubations with bicarbonate. For incubations without bicarbonate they were calculated as the ratios of phosphocreatine formed and total oxygen used from the results in Fig. 2. A comparison of the measured pH gradients in Fig.4A with the P/O ratios in Fig.4B showed that these two quantities are inversely related. The overall P/O ratios were lowered upon increasing the pH gradients and vice versa. However, the relation between the pH gradient and the P/O ratios was not so close as the one observed between pH gradient and energy dissipation. For example in the incubations
+

2 0.5 c
I_

c 3

5 . . . , 0.4
I ,
a 0.3
a l

n ,m

0.2

0.1

8
1.25

0 1.00 ._

.
L

a 0.75

0.50

I I 1 0 [ Creatine] ( rnM)

20

Fig. 4. Transniembrune p H xradients ( A ) and overall P I 0 rurios i B ) qf incubated mitochondria. For the determination of the transmembrane pH gradients in (A) mitochondria from 0.2 g of liver containing 7 mg protein were incubated during 10 min in 1.2 ml of the reaction mixture as described under Methods. For the deter0 1 pCi mination of the sucrose-inaccessible space 50 pCi [ 3 H ] ~plus ['4C]sucrose and for the measurement of the intramitochondrial pH 50 pCi [3H]20 plus 1 pCi 5,5-['4C]dimethyl-2,4-oxazolylidinedione were added to the reaction mixture. Separation o f the mitochondria from the reaction mixture and radioactive assays are described under Methods. The transmembrane pH gradients were calculated as the difference of the measured final pH of the incubations (pH 7.6 for incubations with and pH 6.9 for incubations without bicarbonate) and the intramitochondrial pH. The points given are the mean values of triplicate incubations with mitochondria from the same preparation. The vertical bars denote the largest deviations from the mean values. The same effects were observed in three other experiments. The overall P/O ratios in (B) were calcu~dted as noted in the text. The points given are the mean values of duplicate incubations with mitochondria from the same preparation. The scatter of the results is denoted by the vertical bars. The same effects were observed in four other experiments.

with bicarbonate the pH gradients in the control and in the presence of 10 mM creatine were the same. In contrast, the P/O ratio of the control was well below that measured at 10 mM creatine. This discrepancy may be explained on the basis of earlier experiments. Under the experimental conditions used, i.e. by addition of 4 mM ATP to the medium, a considerable portion of the ATP is split by extramitochondrial

560

Efficiency of Oxidative Phosphorylation

ATPases. Since in the control the extramitochondrial ATP concentration as well as the extramitochondrial ATP/ATP ratio are higher than in the presence of 10 mM creatine, the total amount of ATP split in the medium by these ATPases is also higher in the control than in the presence of creatine [14]. Hence, the overall P/O ratio can be expected to be lower in the control than in the presence of creatine. It is not possible to quantify the contribution of the extramitochondrial ATPases to the overall P/O ratio but in the control incubation. This effect can only be assessed by inhibiting the adeninenucleotidetranslocase, e.g. with atractylate [l]. This inhibition would, however, no longer permit changing the metabolic state of the mitochondria by variation of the creatine concentration in the medium. But still the overall changes of the pH gradients and the P/O ratios in Fig.4 show that the P/O ratio is influenced by the magnitude of the pH gradient. Since, as is shown in this figure, an increase of the pH gradient was always accompanied by a lowering of the P/O ratio, the best overall efficiency of oxidative phosphorylation has to be expected at minimal pH gradients. This conclusion would be in line with the assumption that the H + ion recycling across the inner membrane is an important energy-dissipating process. DISCUSSION Theoretical calculations have shown that the efficiency of oxidative phosphorylation becomes zero when the mitochondria reach the static head or state of maximal energy charge at high extramitochondrial phosphate potentials [20,21]. In many experiments it was shown that the flux ratio of phosphorylation and oxidation, i.e. the P/O ratio of the mitochondria, strongly decreases when mitochondria approach state 4 conditions [1,22-251. Although for the calculation of the efficiency of oxidative phosphorylation in addition to the flux ratio the force ratio of the affinities of phosphorylation and oxidation should be known [20], the experimentally determined P/O ratios demonstrated qualitatively a low efficiency of the oxidative phosphorylation system near state 4 conditions. Similar measurements have been made for calcium transport in mitochondria which is driven by the same high-energy intermediate or high-energy state as oxidative phosphorylation. Again it was found that the Ca/O ratio decreased when mitochondria approach state 4 [2,23]. One can expect the same relation for every other energy-requiring reaction which derives its energy from the same source in the mitochondria as the above-mentioned processes. As was pointed out by Ernster this behaviour of the efficiency of oxidative phosphorylation has important concequences for the calculation of the energy yield of the mitochondria1

phosphorylation system [25]. In addition, since the mitochondria incubated under state 4 conditions are an open system near static head, it seems also very questionable whether the energetics of this system can be analyzed within the framework of classical equilibrium thermodynamics which is strictly applicable only to closed systems at rest. It must be realized that although some components of the phosphorylating machinery show local reversibility, and hence seem to be near equilibrium, the behaviour of an open and a closed system are very different. Therefore, the analysis of the behaviour of the open system with classical thermodynamics under the assumption of equilibrium may lead generally to very misleading conclusions and concepts. The poor phosphorylating efficiency of mitochondria near state 4 has been attributed to the competition of energy-dissipating processes and oxidative phosphorylation for the energy produced in the respiratory chain [25]. These energy-dissipating processes lead to a non-vanishing contribution of the internal entropy production of the system. Such a situation also unequivocally demonstrates the non-equilibrium behaviour of the mitochondria near state 4 [26,27]. Part of these energy-dissipating processes have been identified so far as extramitochondrial ATPase activities [l], calcium recycling across the inner membrane [2], and probably the nicotinamide nucleotide transhydrogenase reaction [25]. In addition, the energy-requiring translocation of ATP and ADP across the inner membrane which is electrical in nature imposes a dependence of the P/O ratio on the extramitochondrial ATP/ADP ratio as was discussed by Klingenberg [13,24]. The main result of this study is the observation of a positive correlation between the magnitude of the transmembrane pH gradient and the total energy dissipation by the mitochondria. Therefore the H + recycling across the inner membrane which consists of a passive leakage of Hions into the matrix and of an energy-utilizing pumping of these ions out of the matrix space, may be another one of the energy-dissipating processes. The slow reentry of Hions into the matrix after an oxygen pulse induced expulsion of these ions into the medium has clearly established the leakiness of the inner membrane for Hions [3,4]. In a recent study Nichols has shown that the H + conductance of the inner membrane is constant up to a protonmotive force of 200 mV [5]. Therefore a linear dependence of the proton leakage on the protonmotive force up to this critical potential can be expected. In order to overcome the problems involved in the use of valinomycin for the determination of the transmembrane electrical potential difference of H + ions across the inner membrane in the presence of high K + concentrations in this study only the dpH across the inner membrane was measured (see Results). It is

J. W. Stucki

561

known that the contribution of the ApH to the total protonmotive force is dependent on the K concentrations of the medium [5]. Since the Kf concentration was constant in our experiments it can be assumed that the observed changes in the ApH also reflect changes in the protonmotive force. Other studies in which the efficiency of oxidative phosphorylation at steady states between state 4 and state 3 was measured were carried out with limiting amounts of hexokinase and glucose as a phosphorylacceptor [24,25]. In the present work the extramitochondrial ATP/ADP was maintained at a steady state with creatine as a phosphoroyl acceptor and with nonlimiting amounts of creatine kinase. The advantage of this method is that a defined extramitochondrial ATP/ADP ratio can be established already at the beginning of the experiment whereas with the hexokinase method this ratio is gradually established only during the incubation. This latter method gives a steady state resulting from the velocities of glucosephosphate formation and oxidative phosphorylation. The disadvantage of the creatine kinase method is that it is not possible to obtain very low ATP/ADP ratios in the medium. This is due to the equilibrium of the reaction and the poor solubility of the creatine. Both the hexokinase and the creatine kinase methods suffer from the fact that the added phosphate is constantly used up for glucose 6-phosphate or creatine phosphate formation so that the phosphate potential is changing during the experiment. This drawback could be overcome by the utilization of ATPase isolated from mitochondria [28]. However, the ATPase method does not allow the direct measurement of the phosphorylated end products like glucose 6-phosphate or creatine phosphate. Such a measurement is mandatory for the present energy balance studies. It has been shown that at elevated ATP concentrations the ATP/ADP ratio in the medium regulates the rate of oxidative phosphorylation since under these conditions the adenine nucleotide translocase becomes the rate-limiting step for this process [28,29]. This conclusion has been critizised by others who found that under special conditions the rate of oxidative phosphorylation was regulated by a phosphate potential which was calculated from the total, intramitochondria1 plus extramitochondrial adenine nucleotides and total phosphate [30,31]. Until it is clearly established which factor regulates the rate of oxidative phosphorylation it is rather difficult to assess the possible influence of the non-steady state phosphate potential obtained with the hexokinase or creatine kinase methods on the overall energetics of oxidative phosphorylation. The rate of pyruvate carboxylation, which is known to be critically dependent on the metabolic state of the mitochondria, was essentially constant during a 10-min incubation also in the pres+

ence of creatine [14]. This indicates that the influence of the non-steady state phosphate potential on energyutilizing reactions in mitochondria is probably minimal under the conditions used.
This work was supported by the Swiss National Sciencc Foundation (Grant no. 3.051.73). The author wishes to thank Professors E. M . Klingenberg and H. Rcuter as wcll as to Dr H. Porzig for criticism and helpful discussions. The expert technical assistance of Mrs U. Michel and Miss L. Lehmann is gratefully acknowledged.

REFERENCES
1. Stucki, J . W. & Walter, P. (1972) Eur. J . Biochcm. 30, 60-72. 2. Stucki. J. W. & Ineichen, E. A. (1974) Eiir. J. Bioclzmz. 48, 365- 375. 3. Mitchell, P. & Moyle, J . (1967) Biorhem. J. 104, 5%-600. 4. Greville, G. D . (1969) in Currenr Topics it7 Biounrrgc~tics(Sanadi, D. R., ed.) vol. 3, pp. 1- 78, Academic Press, New York. 5. Nicholls, D. G. (1974) Eur. J . Biochem. SO, 305-315. 6. Stucki, J. W. (1975) E-xperirntiu iBu.wl) 31, 124. 7. Johnson, D. & Lardy, II. A. (1967) M e t h ~ d .Enzymol. ~ 10, 94 - 96. 8. Graven, S. N., Lardy, H . A. & Rutter, A. (1966) Bioc/?enzi.vlrj~. 5 , 1735-1742. 9. Lamprecht, W., Stein, Ph., Heinz, F. & Weisser, H. (1970) in Metl7oden der enzymati.schen Analyse (Bergmeyer, 14. IJ. ed.) vol. 2, pp. 1729- 1733, Verlag Chemie, Weinheitn/Bergsti-. 10. Pfaff, E. (1965) Inaugural-Dissertation, Marburg. 11. Addanki, S., Cahill, F. D . &, Sotos, J . F. (1968) J . B i d . C/7cm. 243, 2337 2348. 12. Bottger, I., Wieland, O., Brdiczka, D. & Pette, D. (1969) Eur. J . Biochem. 8, 113 - 119. in Biochemistry (Campbell, 13. Klingenberg, M. (1 970) in ESSUJ;.~ P. N. & Dickens, F., eds) vol. 6, pp. 119-159, Academic Press, London. 14. Stucki, J. W., Brawand, F. & Walter, P. (1972) Eur. J . Biochem. 27,1X1-191. 15. Walter, P., Paetkau, V. & Lardy, H. A. (1966) J. Riol. Chem. 241, 2523-2532. 16. Lopes-Cardozo, M., Vaartjes, W. J. & van den Bergh, S. G . (1972) FEBS Lett. 28, 265-270. 17. Rottenberg, H. (1970) Eur. J. Biochem. 15, 22-28. 18. Padan, E. & Rottenberg, H. (1973) Eur, J . Biochem. 40, 431 437. 19. Mitchell, P. &, Moyle, J . (1969) Eur. J . Biod7em. 7 , 471 -484. 20. Kedem, 0. & Caplan, S. R. (1965) Trans. Faradu.v .So(,. 61, 1897 - 1911. 21. Rottenberg, H., Caplan, S. R . & Essig, A. (1970) in Men7hranes and Ion Transport (Bittar, E. E., ed.) vol. 1, pp. 165191, Wiley-Interscience, London. 22. Lardy, H. A. 81 Wellman, H. (1952) J. B i d . Chern. 195, 215224. 23. Carafoli, E., Rossi, C. S. & Lehninger, A. L. (1965) Biochcm. Biophjs. Res. Commun. 19, 609 - 614. 24. Klingcnherg, M. (1972) in Mitochondria: Biogmesis and Bioenergetics (van den Bergh, S. G., Rorst, P., van Deenen, L. L. M., Riemersma, J. C., Slater, E. C. & Tager, J . M., eds) pp. 147-162, North Holland, Amsterdam. 25 Ernster, L. & Nordenbrand, K. (1974) in Dynamics qf EnergyTrunsducing Memhrunes (Ernster, L., Estabrook, R. W. & Slater, E. C., eds) B.B.A. Library, vol. 13, pp. 283-288, Elsevier Scientific Publishing Company, Amsterdam.
-

562
26. Prigogine, I. (1967) Introduction to Thermodynamics of h e versible Processes. Interscience Publishers, New York. 27. Katchalsky, A. & Curran, P. F. (1965) Nonequilibriuni Theumodynamics in Biophysics, Harvard University Press, Cambridge, Massachusetts. 25. Davis, E. J. & Lumeng, L. (1975) J . Biol. Chem. 250, 22752282.

J. W. Stucki: Efficiency of Oxidative Phosphorylation

29. Slater, E. C., Rosing, J. & Mol, A. (1973) Biochim. Biophys. Acta, 292, 534- 553. 30. Owen, Ch. S. & Wilson, 0. F. (1974) Arch. Biochem. Biophys. 161, 581 -591. 31. Erecinska, M., Veech, R. L. & Wilson, D. F. (1974) Arch. Biochem. Biophys. 160,412- 421.

J . W. Stucki, Service de Chimie Physique 11, Faculte des Sciences de lU.L.B., Code Postal 231, Campus Plaine de IU.L.B., Boulevard du Triomphe, B-1050 Bruxelles, Belgium