Studies on the relationship between ketogenesis and pyruvate oxidation in isolated rat liver mitochondria.

The regulation of the pyruvate dehydrogenase multienzyme complex was studied in the perfused rat liver and in isolated rat liver mitochondria. The rate of “CO, production from added [IJClpyruvate was utilized as a monitor of the pyruvate dehydrogenase activity in both preparations. In the perfused liver infusion of the medium chain length fatty acid octanoate resulted in an acceleration of pyruvate decarboxylation at low (0.05 mu) perfusate pyruvate concentrations, while an inhibition of “CO, production was observed at high (5.0 rn>l) pyruvate concentrations. Mitochondrial experiments were performed to define metabolic conditions which mimic the differential effect of fatty acid on the flux through the pyruvate dehydrogenase reaction. A stimulation of pyruvate decarboxylation by fatty acid was observed in mitochondrial incubations maintained in State 4 in the absence of a source of oxalacetate, r,-malate, but only at low pyruvate concentrations. Changes in the intramitochondrial ATPlADP ratio likely were not involved in this regulatory process as there occurred no obligatory inverse relationship between the activity state of pyruvate dehydrogenase and the ATP/ADP ratio. Measurement of pyruvate dehydrogenase activity in the mitochondrial incubations indic%ed that the increase in the rate of pyruvate decarboxylation caused by octanoate addition was due at least in part to an interconversion of the pyruvate dehydrogenase complex to its active form. The stimulation of flux through pyruvate dehydrogenase and the interconversion of the enzyme complex to its active form by fatty acid in rat liver mitochondria likely involves a relationship between the process of ketogenesis and the regulation of pyruvate dehydrogenase by the substrateleffector pyruvate. Consistent with our previous proposal (based on experimental results with the isolated perfused rat liver) it is suggested that the rapid efflux of intramitochondrially generated acetoacetate resulting from fatty acid oxidation causes an acceleration of pyruvate entry into the mitochondrial matrix via the monocarboxylate trans-

Since this system of interconversion of pyruvate dehydrogenase between its active and inactive forms, besides being regulated by adenine nucleotides (8ll), pyruvate (8,(12)(13)(14), and mono-and divalent metal cations (6,(13)(14)(15)(16)(17)(18), is subject to modulation by the oxidationreduction state of the mitochondrial nicotinamide adenine dinucleotides (e.g. NADH/NAD+) and the acetyl-CoA/CoASH ratio (18-221, both types of control mechanisms (e.g. feedback inhibition by products and interconversion of active and inactive pyruvate dehydrogenase) have been implicated in the frequently observed inhibition of pyruvate oxidation during the oxidation of fatty acids. This inhibitory effect of fatty acids has been observed in perfused organs (8,(23)(24)(25)(26)(27) and in isolated mitochondria (11, 19-22, 28, 29)  were incubated as before at 37" for 5 min at which time 0.45-ml samples were withdrawn and carefully layered on 0.5 ml of silicone oil (specific gravity 1.054) above 0.1 ml of 1.5 N perchloric acid in a microcentrifuge tube (1.5 ml) and spun for 1 min at top speed in an Eppendorf microcentrifuge. Samples from both above (0.02 ml) and below (0.05 ml) the oil were counted for 14C and 3H emissions in 10 ml of Aquasol scintillation mixture.
In all experiments the sucrose-accessible space and, by difference from the mitochondrial water space, the matrix space were determined using [U-'Clsucrose (2.5 @Zi) in place of substrate, both in the presence and absence of octanoate (2 mM

RESULTS
Most studies performed in complex metabolic systems concerning the effect of fatty acid oxidation on the regulation of pyruvate dehydrogenase have employed pyruvate concentrations in the 1 to 10 mM range. These investigations have consistently demonstrated that fatty acids lead to a restricted flux of carbon through the pyruvate dehydrogenase reaction largely due to a conversion of the multienzyme complex to its inactive, phosphorylated form. The observation of Scholz et al. (30) that long chain (oleate) or medium chain (octanoate) fatty acids or P-hydroxybutyrate caused up to a 3-fold enhancement of pyruvate dehydrogenase flux in the perfused rat liver at low perfusate pyruvate concentrations was the first indication that fatty acid oxidation may cause other than an inhibition or inactivation of the enzyme complex. As a starting point in an attempt to develop experimental support for a plausible mechanism for this fatty acid-mediated enhancement of pyruvate dehydrogenase flux in the liver, the experiment shown in Fig. 1   seemed to be the most promising.
As can be seen in Table I  A time course of the octanoate-mediated stimulation of pyruvate dehydrogenase flux (data not shown) indicated that the enhancement of pyruvate decarboxylation activity in the mitochondrial suspension increased with more prolonged incubation times. Incubation periods in the experiment depicted in Fig. 2 were limited by the fact that the mitochondria were maintained in State 3 which in prolonged incubations at low pyruvate concentrations would have resulted in a depletion of the substrate. With mitochondria incubated in State 4 the much greater stimulation in the rate of pyruvate decarboxylation by octanoate when the incubations were continued for a longer time period is demonstrated in Fig. 3.
Outlined in Table II are the results of an experiment designed to determine whether the changes in mitochondrial pyruvate flux produced by octanoate metabolism were reflected in altered activation states of the isolated pyruvate dehydrogenase complex. The results of this series of experiments indicated that in mitochondria maintained in State 4 conditions octanoate metabolism affected the activity of subsequently extracted pyruvate dehydrogenase in a manner entirely consistent with the previously observed actions (Figs. 2 and 3) of octanoate on pyruvate decarboxylation in intact mitochondria.
Again, only at low pyruvate concentrations (0.1 mM) and in the absence of L-malate was a significant increase in pyruvate dehydrogenase activity in the presence of octanoate observed. Furthermore, since samples of incubation prior to pyruvate dehydrogenase extraction and assay were diluted 4-fold to eliminate complications from end product inhibition, it may be assumed that octanoate inhibition at high pyruvate concentrations and stimulation in the absence of L-malate at low pyruvate concentrations of isolated pyruvate dehydrogenase activity reflect alterations by covalent modification in the activation state of this multienzyme complex. Thus, in the light of this data and the Michaelis-Menten kinetic analysis discussed above, it seems reasonable to suggest that the enhanced pyruvate flux seen under certain metabolic conditions during octanoate metabolism likely involves the inter-   Table II, it is apparent that rat liver mitochondria, incubated in State 4 in the absence of malate and at low medium pyruvate concentrations, represent a suitable metabolic situation in which to investigate possible mechanisms for this stimulatory effect of fatty acid oxidation on the activation state of the pyruvate dehydrogenase multienzyme complex in this mitochondrial system. In perfused rat livers to which oleate (56, 57) or octanoate (46) were infused and in uiuo during states of accelerated fat oxidation such as fasting or diabetes (58, 591, a decrease in the ratio of ATP/ADP has been reported. Since a direct correlation of the ATPIADP ratio with the pyruvate dehydrogenase kinase activity and thus an inverse relationship between the ATPIADP ratio and the pyruvate dehydrogenase activity have been indicated frequently (8-11, 14, 281, a decrease in the ATP/ADP ratio would have the effect of stimulating the pyruvate dehydrogenase activity in this mitochondrial system. However, as shown in Table III, the effects of intramitochondrial octanoate activation (60) and metabolism on the adenine nucleotide levels in isolated liver mitochondria incubated in State 4 indicated changes in the ATPIADP ratio which were inconsistent with the enhancement by octanoate of pyruvate dehydrogenase activity being mediated via this parameter. Although at both high and low pyruvate concen- trations in the presence of L-malate, octanoate addition resulted in a reduction in the ATP/ADP ratio, under these incubation conditions there occurred no marked stimulation of pyruvate decarboxylation by octanoate (Fig. 2, Table IV). Conversely, in the absence of L-malate at low pyruvate concentrations where an octanoate-mediated stimulation of pyruvate decarboxylation was observed (Figs. 2 and 3, Table IV) the addition of octanoate was associated with a markedly elevated ATPIADP ratio which again is not consistent with an ATP/ ADP-mediated activation of pyruvate dehydrogenase via the kinase-phosphatase system. Hence, it is evident that another regulatory factor must be primary in this enhancement of pyruvate dehydrogenase activity by fatty acid. Of the end products of ,6 oxidation of fatty acids NADH and acetyl-CoA have demonstrated effects as activators of the pyruvate dehydrogenase kinase. Elevation of these two species upon initiation of rapid fatty acid oxidation should result in an inhibition of pyruvate decarboxylation in this system (see introduction to the text). In liver a frequent consequence of the initiation of rapid /3 oxidation of fatty acids is the synthesis of /3-hydroxybutyrate and acetoacetate. While no direct effects of the ketone bodies per se have been demonstrated on the isolated pyruvate dehydrogenase multienzyme complex or its two regulatory enzymes, acetoacetate has been shown by Papa and Paradies (38) to be an excellent anionic exchange species for pyruvate on the monocarboxylate exchanger in the inner mitochondrial membrane of liver mitochondria. Therefore, it seems plausible that under conditions where pyruvate entry into the mitochondria might be limiting, e.g. at low medium pyruvate concentrations, any process which could accelerate the exchange or translocation of pyruvate across the membrane would lead to an enhancement of the pyruvate dehydrogenase flux through two mechanisms: (al through increased substrate (pyruvatel supply, and (b) through the documented ability of pyruvate to inhibit the pyruvate dehydrogenase kinase with subsequent activation of the multienzyme complex (8,(12)(13)(14). In either event or both together, initiation of rapid synthesis of intramitochondrial which could exchange for extramitochondrial pyruvate would represent an intriguing possibility to explain the fatty acid-mediated increase in the pyruvate dehydrogenase activity.
In Table IV the results of an experiment are presented in which the rates of ketogenesis and pyruvate dehydrogenase are compared in the presence and absence of octanoate under a variety of experimental conditions. A number of observations are pertinent in this experiment.
First, the degree of reduction of the P-hydroxybutyratelacetoacetate ratio produced in the presence of octanoate should be compared with those in the liver perfusion experiment shown in Fig. 1. Only the ratios of P-hydroxybutyratelacetoacetate in the mitochondrial incubations in the presence of L-malate in State 4 approached the ratios seen in the perfused liver experiment. This observation, together with the fact that the differential effects of octanoate on the pyruvate decarboxylation seen in the perfused liver could only be stimulated in the mitochondrial system in State 4 (Table IV, Fig. 2), is consistent with the hypothesis (61) that the in uiuo metabolic situation in the liver resembles an ADP-controlled State 4. Second, support for the general assumption (see Ref. 50) that ketogenic rates are dependent upon excess of the rates of acetyl-CoA production over utilization by citrate synthase which in turn is controlled by oxalacetate availability (34) is seen in the diminished rates of ketogenesis in the presence of L-malate. Finally, and of primary interest, is the effect of Lmalate on the stimulation of pyruvate dehydrogenase activity by octanoate in mitochondria incubated in State 4 at low pyruvate concentrations. When a source of oxalacetate was included in these incubations, the increase in the rates of ketogenesis upon addition of octanoate was negligible. Similarly, the increase in pyruvate dehydrogenase flux, possibly explicable in terms of the data presented in Table III, was also relatively small. In the absence of cmalate, however, the inclusion of octanoate in the incubation resulted in the maximum stimulation in the pyruvate decarboxylation activity and associated with this stimulation of pyruvate dehydrogenase was a large increase in acetoacetate production. Hence, a primary conclusion from this experiment may be that the correlation between the increased acetoacetate production and the increased pyruvate dehydrogenase activity resulting from octanoate addition, although not absolute proof, is certainly consistent with the suggestion that the elevated ketone body eMux promoting pyruvate accumulation in the mitochondrial compartment could be the factor responsible for the increased pyruvate dehydrogenase activity.
Further support for this suggestion was obtained by investigating the effect of initiation of rapid ketogenic rates from octanoate on pyruvate accumulation in isolated liver mitochondria (see Fig. 4). It can be seen that there occurred a significant accumulation of pyruvate by the mitochondria and this accumulation was enhanced in the presence of octanoate is compared to the control in which octanoate was omitted. The omission of succinate, added as an energy source in the control incubations, which should accelerate fatty acid oxidation (62) and decrease the reduction state of the ketone bodies produced (Fig. 4) (Tables  II and IV), there was no consistent obligatory inverse relationship between the activity state of the pyruvate dehydrogenase complex and the intramitochondrial ATP/ADP ratio (Table III)