Sodium pyruvate, oxaloacetic acid, L-dipalmitoyl phosphatidylcholine (DPPC), Coenzyme Q₁₀ (CoQ), hydrogen peroxide, and L-malic acid were purchased from Wako Pure Chemical Industry, NAD⁺, NADH and adenosine diphosphate from Oriental Yeast Co., Ltd. and pyruvate dehydrogenase, Coenzyme A (CoA), citrate synthase, aconitase, cytochrome c, cytochrome c reductase, cytochrome c oxidase, luciferin, and luciferase were purchased from Sigma-Aldrich Co. Fresh beef (fillet part) was obtained from a butcher.

Mitochondria were prepared from 2 g of beef using a mitochondrial isolation kit for tissues and cultured cells of BioChain Institute Inc. After completing all the steps according to the kit, the mitochondria were stained with Janus Green. As a result, it was found that mitochondria could be isolated because they were stained blue.

Liposome solution was prepared as shown preceding paper [11]. At first, toluene containing 1 mM DPPC and 10 mM CoQ was layered on the buffer solution of equal volume.

The transfer of DPPC was confirmed with the following method. Two days after laying the toluene solution of DPPC without CoQ on the aqueous solution, the toluene solution of DPPC was evaporated to dryness and the mass of DPPC remained was weighted. The result indicated that 86.4% of DPPC transferred into the aqueous solution from toluene solution. Likewise, concentration of remained CoQ in the toluene solution was determined by measuring the absorbance at 281 nm. From this result, it was found that the concentration of CoQ transferred into the aqueous solution was 1.54 mM out of 10 mM.

The fluorescence intensity was measured at 30˚C using a spectroscopic fluorometer (Shimadzu RF-5300). The sample volume for each measurement was 2 mL The fluorescence of NADH was measured at 340 nm (exciting light) and 460 nm (emission light). To measure the fluorescence of ATP, 1 mM luciferin (0.1 mL), 1 mg/mL luciferase (0.1 mL), and 1 mM Mg²⁺ ion (0.05 mL) was added to a solution. Fluorescence was measured at 550 nm.

In order to measure the pH within mitochondria, we used a pH-dependent fluorescent reagent, BCECF-AF, (2’, 7’-bis-(2-carboxyethyl)-5-(and -6) -carboxyfluorescein, acetoxymethyl ester) [13]. After incubating the mitochondrial solution in the presence of 0.1 mM BCECF-AF, the mitochondria containing BCECF-AF were separated from the external solution by centrifugation for 30 minutes. The fluorescence intensity of BCECF was measured at 500 nm (excitation light) and 530 nm (emission light).

The time course of the absorbance was measured by spectrophotometer (Shimadzu UV-1800). The absorbance when using a semipermeable membrane was measured by as follows. Apparatus for measuring the absorbance is shown in Figure 1. Cuvette was mounted in a brass block which was thermally controlled by circulating water at constant temperature, 37˚C. Glass tube covered with dialysis membrane at the one end was inserted into a cuvette which contained enzyme solution (2 mL). Substrate solution (1 mL) was put into the glass tube.

First, we investigated NADH oscillations when only pyruvate entered the mitochondria. When pyruvate crosses the inner mitochondrial membrane, the mitochondrial pyruvate carrier (MPC), a protein embedded in the membrane, promotes pyruvate to cross the membrane [14] [15]. After the reaction of pyruvate with NAD⁺ and CoA in the inner mitochondrial membrane, the citric acid cycle and respiratory chain reaction proceed. In order to confirm the presence of NADH oscillations in these processes, the isolated mitochondria were dispersed in Tris buffer at pH 8.0 in which pyruvate was dissolved, and the absorbance of NADH was measured at 340 nm. With 0.2 mM pyruvate, small amplitude

oscillations occurred in short cycles. When the concentration of pyruvate was increased, oscillations with large amplitude and long period were observed after a long period of time (Figure 2(a)). It was demonstrated that the gradual entry of pyruvate into the mitochondria would cause NADH oscillations.

Next, ADP, a substrate for ATP synthase, was added to the solution containing mitochondria and pyruvate. Also in this case, the time course of absorbance by NADH was also measured. As shown in Figure 2(b), NADH oscillation occurred as well as without ADP. The oscillation pattern was the same as without ADP.

No ATP oscillations were observed under the same conditions as in Figure 2(a). However, when ADP was supplied, short-period oscillations (10 - 20 minutes) of ATP were observed in the concentration range of pyruvate 20 mM or less at a fixed concentration of ADP (6.7 mM) after the induction period (Figure 3(a)). Figure 3(b) shows a phase diagram showing the concentration range of pyruvate and ADP in which the oscillation occurs.

NADH produced by glycolysis is also used in the respiratory chain. NADH cannot

permeate the membrane, but NAD⁺ can be converted to NADH using the malate-aspartate shuttle [16] [17]. Therefore, by measuring the fluorescence by NADH, we investigated the oscillation when mitochondria were added to a solution containing both NAD⁺ and malate.

It was found that NADH oscillation occurs even with the membrane permeation of malate in the presence of NAD⁺ without the supply of ADP (Figure 4(a)). However, in this case, NADH oscillation ends within 6 hours. In the actual malate-aspartate shuttle, in addition to the reaction between malate and NAD⁺, the reaction between the product oxaloacetate and glutamate takes place. Glutamic acid was not added in this study. Therefore, NADH formation may be suppressed. However, when ADP is added, NADH oscillation continues for a long time. In the reaction up to Complex IV, the formation of NADH ends early without ADP, but the reaction continues with the addition of ATP, and it is thought that oscillation is likely to occur even without glutamate. NADH oscillations were shown to occur in a wide range of NAD⁺ and malate, both with and without ADP supply. As shown in Figure 4(b) and Figure 4(c), with the addition of ADP, both NADH and ATP oscillations were observed in the long term

with small amplitudes. It was found that oscillation occurs without going through the process of the citric acid cycle.

Since H⁺ is produced by the reaction of both respiratory chain Complexes I and III, a pH-dependent fluorescent reagent was used to determine the pH in the mitochondria. We investigated the time course of mitochondrial pH in both pyruvate membrane permeation and NAD⁺ and malate permeation systems. The results are shown in Figure 5. In the former, the H⁺ concentration increased with H⁺ oscillation. On the other hand, in the latter case, H⁺ oscillated at first, but after a while the H⁺ concentration decreased, and the oscillation became weaker. The cause of the decrease may be due to the absence of the required glutamate addition during the malate-aspartate shuttle.

In addition to the investigation with mitochondria, in order to verify an assumption that NADH and the other intermediates oscillations triggered by the gradual entry of first substrates (pyruvate or NADH) occur continuously in the citric acid cycle and the respiratory chain, we carried out the following experiments by using dialysis membrane.

In addition to NADH produced by the citric acid cycle, NADH produced by glycolysis is also utilized in the respiratory chain. Although NADH is not able to permeate through mitochondrial membrane, NADH is produced from NAD⁺ within matrix of mitochondria by the malate-aspartate shuttle [16]. In order to prove an assumption that the gradual entry of NADH caused oscillatory reactions of intermediates existing in respiratory chain, a method using dialysis membrane was employed as a simple model of permeation through membrane.

At first, the direct reaction between NADH and coenzyme Q (CoQ) was investigated. NADH solution was put in upper phase of apparatus of Figure 1, while liposome solution of L-dipalmitoyl phosphatidylcholine (DPPC) containing CoA was put in lower phase. Time course of the absorbance at 340 nm, which is assigned to the absorption of NADH, was measured. As shown in Figure 6(a), direct reaction of NADH and CoQ occurred after an increase in the concentration of NADH accompanying the permeation of NADH across the

membrane. Then the repetition of increase and decrease of absorbance occurred, i.e., oscillation was observed. The decrease in pyruvate concentration led to the oscillation with short period and small amplitude after that of long period and large amplitude.

Furthermore, we added cytochrome c and cytochrome c reductase to lower phase. The reaction in lower phase is thought to be equivalent to that of Complex III in respiratory chain. In this reaction, the electron of NADH is transferred to cytochrome c via CoQ. As the wavelength of the absorption of NADH and of reduced type of cytochrome c is 340 nm and 550 nm, respectively, the absorptions at 340 nm and 550 nm were measured at the same time. As shown in Figure 6(b), the oscillation of NADH appeared explicitly. Correspond to the oscillation of NADH, the oscillation of cytochrome c with same period was observed. Two oscillations were confirmed to be synchronized.

Finally, small quantity of cytochrome c oxidase was also added to lower phase. This reaction is thought of corresponding to that of Complex IV in respiratory chain. As shown in Figure 6(c), the addition of cytochrome c oxidase led to little change of oscillation pattern at the same concentration as cytochrome c reductase. So, we reduced the concentration of cytochrome c reductase. As shown in Figure 6(d), when the concentration of cytochrome c oxidase equals to that of cytochrome c reductase, the oscillation with period of about 20 h occurred in both of NADH and cytochrome c. In greater quantity of cytochrome c oxidase than cytochrome c reductase, distinct oscillation with short period appeared after induction period. We found that the oscillatory reactions of NADH and CoQ induced continuous oscillatory reactions for the respiratory chain.

Reactions using semipermeable membranes caused oscillations with longer period than mitochondrial. It is considered that this is because the oxygen supply is blocked by the presence of the upper phase sandwiched between the dialysis membranes, and the oxygen supply from the air required for the cytochrome c oxidase reaction is insufficient. To supply oxygen in the system of Figure 6(d), hydrogen peroxide was added to the upper phase and catalase was added to the lower phase. Under this condition, the oscillation of oxygen is not found to occur, and oxygen is gradually generated [7]. It was found that short-period NADH oscillation appears due to the supply of oxygen without going through the induction period (Figure 6(e)). Further increases in the concentration of cytochrome c reductase will further shorten the oscillation period. Oxygen supply seems to accelerate the reaction of cytochrome c oxidase and facilitate the oscillation.

Pyruvate produced by glycolysis flows through mitochondrial membrane and is changed to acetyl-coenzyme A (CoA) by pyruvate dehydrogenase which catalyzes the reaction of pyruvate, NAD⁺ and CoA. When pyruvate passes across the

mitochondrial inner membrane, mitochondrial pyruvate carrier (MPC), which is protein embedded within membrane, facilitates the pass of pyruvate across membrane [14] [15]. Since the flow of pyruvate is also thought to be caused by the difference in electrochemical potential between the inside and outside of the mitochondrial inner membrane, it can be assumed that a semipermeable membrane can be used instead of the mitochondrial membrane for simplicity.

At first, in apparatus of Figure 1, we put pyruvate solution in upper phase, while pyruvate dehydrogenase, CoA and NAD⁺ in lower phase, for measuring the formation of acetyl-CoA and NADH. Time course of the absorbance at 340nm due to the absorption of NADH was measured.

As shown in Figure 7(a), when the concentration of pyruvate was increased at the fixed concentrations of NAD⁺ and CoA, the period of NADH oscillation became short. On the other hand, decreasing the concentration of CoA at the fixed concentration of pyruvate and NAD⁺ led to a stepwise increase in absorbance (Figure 7(b)). From these results, it was considered that the oscillation of acetyl-CoA, which was another product of this reaction, also happened.

As it was found that the formation of acetyl-CoA was oscillatory reaction in the mediation of semipermeable membrane, we conceived those subsequent reactions in citric acid cycle oscillatesimilarly. Therefore, oxaloacetic acid, citrate synthase and aconitase were also added to lower phase. The absorbance of at both 340 nm and at 240 nm, which were due to the absorptions of NADH and cis-aconitic acid, respectively, were measured at the same time.

As shown in Figure 7(c), not only the oscillation of NADH, but also that of cis-aconitic acid, the intermediate in the reaction of aconitase, was observed. When pyruvate concentration was small, the oscillations of NADH and cis-aconitic acid were restrained. However, as pyruvate concentration was increased, two oscillations were synchronized and the amplitude of cis-aconitic acid oscillation became large. Further increase of pyruvate led to short period and small amplitude oscillation of NADH. The oscillation of NADH is thought to induce the oscillatory reaction of aconitase via that of citrate synthase. The reactions of citric synthase and of aconitase are those in first stage of citric acid cycle. As cis-aconitic acid is the intermediate of citric acid cycle, this result suggests that the oscillatory reaction will be passed down to remained reactions in citric acid cycle.