Regulation of Hepatic Gluconeogenesis in the Guinea Pig by Fatty Acids and Ammonia*

Octanoate and L-palmitylcarnitine inhibited the synthesis of P-enolpyruvate from a-ketoglutarate and malate by isolated guinea pig liver mitochondria. A 50% reduction in P-enolpyruvate formation the a-ketoglutarate the of P-enolpyruvate and the reduction of the to P-hydroxybutyrate reversed the the P-enolpyruvate

Octanoate and L-palmitylcarnitine inhibited the synthesis of P-enolpyruvate from a-ketoglutarate and malate by isolated guinea pig liver mitochondria.
A 50% reduction in P-enolpyruvate formation was obtained with 0.1 to 0.2 mM octanoate or with 0.06 to 0.10 mM L-palmitylcarnitine.
At these concentrations, oxidative phosphorylation remained intact and only much higher concentrations of fatty acids altered this process. The addition of NH&l in the presence of malate and increasing concentrations of cu-ketoglutarate (or oice versa) enhanced the formation of glutamate, aspartate, and P-enolpyruvate.
The addition of increasing concentrations of NH&l in the presence of fixed amounts of malate and a-ketoglutarate had a similar effect. Furthermore, the inhibition of P-enolpyruvate synthesis by fatty acids and the reduction of the acetoacetate to P-hydroxybutyrate ratio were reversed by the addition of NH&l. Cycloheximide, which blocks energy transfer at site 1 of the respiratory chain, decreased P-enolpyruvate formation. When cycloheximide and either octanoate or L-palmitylcarnitine were added together, there was an even greater reduction in P-enolpyruvate synthesis from either malate or a-ketoglutarate than was noted with either fatty acid alone. Since cycloheximide lowers the rate of ATP synthesis this may in turn reduce Penolpyruvate formation by a mechanism independent of changes in the mitochondrial NAD+/NADH ratio caused by fatty acids.
In the isolated perfused liver metabolizing lactate, the inhibitory effect of octanoate on gluconeogenesis was partially relieved by the addition of 1 mM NH&l, but remained unchanged in the presence of 2 mM NH&l, despite a highly oxidized NAD+/NADH ratio in the mitochondria. In contrast to glucose synthesis, urea formation was markedly increased during the infusion of 1 mM as well as 2 mM NH&l. After cessation of NH&l infusion, there was an increase in glucose production, to a rate as high as that observed in the absence of octanoate. This increase was accompanied by the disappearance of alanine, aspartate, and glutamate which had been stored in the liver during NH&l infusion. Urea synthesis also decreased progressively. These results indicate that gluconeogenesis in guinea pig liver is regulated, in part, by alterations in the mitochondrial oxidation-reduction state. However, the modulation of this effect by changing the concentrations of intermediates of the aspartate aminotransferase reaction indicates competition for oxalacetate between the aminotransferase reaction and P-enolpyruvate carboxykinase.
Species differences in the regulation of hepatic gluconeogenesis have been well characterized (1,2). A comparison of rat and guinea pig livers has shown that the presence of a mitochondrial P-enolpyruvate carboxykinase in the latter species is partly responsible for a number of differences in hepatic glucose synthesis (3-7). Fatty acid oxidation by guinea pig liver mitochondria shifts the NAD+/NADH ratio toward reduction and markedly reduces the rate of P-enolpyruvate formation (3,4,8 fused guinea pig livers markedly diminishes gluconeogenesis from lactate and alanine (4, 8), but stimulates this process in rat liver (9-12). The inhibitory effect of fatty acids on gluconeogenesis coincides with a decrease in the mitochondrial NAD+/NADH ratio as calculated from the equilibrium of phydroxybutyrate dehydrogenase and glutamate dehydrogenase reactions.
While there is no doubt that fatty acids can inhibit hepatic gluconeogenesis in guinea pig liver in uitro, the physiological significance of this finding has not been established. During fasting the oxidation of fatty acids by the liver is enhanced, yet gluconeogenesis increases (4). Furthermore, the mitochondrial NAD+/NADH ratio in guinea pig livers, freezeclamped in uiuo, shifts markedly toward oxidation after 48 to 72 hours of fasting (13). The relationship between this shift toward oxidation in the mitochondrial oxidation-reduction state and the oxidation of fatty acids has not been clearly delineated. Other factors such as the metabolism of amino acids and of ammonia in the liver might also contribute to changes to the in vioo mitochondrial oxidation-reduction state. In this report we have studied the interaction of ammonia metabolism with gluconeogenesis and ureogenesis in perfused guinea pig liver. Our results indicate that ammonia, at concentrations close to those noted in intact liver, will reverse the inhibitory effect of fatty acids on glucose synthesis from lactate by perfused guinea pig liver.  (Fig. 1). A 50% inhibition of P-enolpyruvate formation from malate was noted at 0.2 mM octanoate, whereas the same degree of inhibition with cY-ketoglutarate as substrate required only 0.1 mM octanoate. This inhibition was not due to an uncoupling of oxidative phosphorylation since the levels of AMP remained constant over the entire concentration range of octanoate added to the incubation medium. Also, at the lowest concentration of octanoate there was a sharp increase in ATP formation when malate was the substrate. This activation was consistent with our observations of oxygen consumption by guinea pig liver mitochondria metabolizing malate plus glutamate (Fig.  2). The lack of increase in AMP formation confirmed the functional stability of these mitochondria at the concentrations of fatty acid used. We also noted ( Fig. 3) a 50% inhibition of P-enolpyruvate formation with approximately 0.08 and 0.06 mM L-palmitylcarnitine in the presence of malate and cr-ketoglutarate, respectively.
The high rate of ATP synthesis and the absence of changes in the concentration of AMP indicated that the mitochondria remained coupled while metabolizing either fatty acid.
Work in our laboratory (7) and by Bryla (28) has indicated that alterations in the mitochondrial oxidation-reduction state during the oxidation of fatty acids is partly responsible for the marked reduction in the rate of mitochondrial P-enolpyruvate synthesis by isolated guinea pig, human, and rabbit liver mitochondria metabolizing a variety of substrates. To test this hypothesis further, we measured the effect of NH ,CI on P-enolpyruvate formation from a-ketoglutarate and malate (Fig. 4). NH&l should cause an oxidation of the mitochondrial oxidation-reduction state by shifting the equilibrium of the glutamate dehydrogenase reaction toward glutamate formation with a resultant net generation of NAD+. Such a shift would be dependent on a source of a-ketoglutarate and should stimulate P-enolpyruvate synthesis. As shown in Fig. 4, NH,Cl in a concentration of up to 2 mM increased P-enolpyruvate forma-  tion of 2 mM and the levels of malate were increased to 4 mM. Fig. 5 also shows that aspartate aminotransferase can compete effectively with P-enolpyruvate carboxykinase since aspartate synthesis also increased markedly. Malate was a better source of carbon for both P-enolpyruvate and aspartate synthesis than was a-ketoglutarate, probably reflecting a more rapid rate of oxalacetate formation from malate by guinea pig liver mitochondria. Also, the synthesis of glutamate was always higher in the presence of octanoate, indicating that NADH formed by fatty acid oxidation was being used (in the presence of NH&l) to drive the glutamate dehydrogenase reaction toward glutamate synthesis.
The relationship between the rate of P-enolpyruvate synthesis and the production of ketone bodies, glutamate, end aspartate by guinea pig liver mitochondria metabolizing 2 mM malate plus 2 mM cY-ketoglutarate is further demonstrated in Table I. Octanoate and Lpalmitylcarnitine both decreased the rate of P-enolpyruvate formation and this effect could be reversed by NH&l.
Cycloheximide, which blocks energy transfer at site 1 of the respiratory chain (29), also decreased P-enolpyruvate synthesis, but when added to mitochondria together with fatty acids, there was an even greater decrease in P-enolpyruvate formation. This effect is probably related to the profound reduction in the oxidation-reduction state of the mitochondria when both cycloheximide and either octanoate or Lpalmitylcarnitine were present together. The ratio of acetoacetate to /3-hydroxybutyrate decreased from 3.7 in control incubations to 0.9 when L-palmitylcarnitine and cycloheximide were combined. This was accompanied by a decrease in ATP synthesis in state 3 when cycloheximide and r.-palmitylcarnitine were present, again consistent with a block of energy transfer at site 1 of the respiratory chain. The addition of NH&l to the mitochondrial incubation medium partly relieved P-enolpyruvate formation in the presence of cycloheximide without affecting the markedly reduced rate of ATP formation.
In those experiments where fatty acids and NH&l were present together, the rate of glutamate, but not aspartate, synthesis could be markedly increased by cycloheximide. Since the level of ATP synthesis remained reduced after cycloheximide addition, it is reasonable to assume that the increased glutamate formation was due to a further increase in NADH generation caused by a block of site 1. P-enolpyruvate synthesis was also decreased, reflecting the reduction in the mitochondrial oxidation state as shown by the decreased acetoacetate to P-hydroxybutyrate ratio (2.5 with cyclo-8981 heximide, NH&l, and octanoate uersus 5.9 with octanoate plus NH&l).
The direct transamination of oxalacetate to aspartate by guinea pig liver mitochondria was markedly inhibited by aminoxyacetic acid, a compound which blocks aspartate aminotransferase (30,31). When mitochondria were metabolizing malate and a-ketoglutarate, NH&l addition resulted in the formation of 76 nmol of aspartate/mg of mitochondrial protein in 10 min (Table I). Aminoxyacetic acid reduced the rate of aspartate synthesis to 6.1 nmol/mg of protein, whereas the synthesis of glutamate and of P-enolpyruvate was not blocked.
The inhibition of aspartate aminotransferase by aminoxyacetic acid caused an increase in P-enolpyruvate synthesis by guinea pig liver mitochondria when NH&l was included in the incubation medium. Thus the rate of P-enolpyruvate formation increased from 125 nmol/mg of protein in 10 min with NH&l present to 151 nmol when aminoxyacetic acid and NH&l were added together. This probably reflected the decreased use of oxalacetate for aspartate synthesis, thereby diverting substrate for the P-enolpyruvate carboxykinase reaction.
There is a good correlation between the control of Penolpyruvate synthesis by guinea pig liver mitochondria and the regulation of gluconeogenesis in perfused guinea pig livers (4, 8, 32). Octanoate infusion sharply decreased glucose synthesis from lactate (Fig. 6) while 1 mM NH&l added together with octanoate reversed the pattern of inhibition and caused a return of glucose synthesis, to levels noted before octanoate infusion. This effect of NH,Cl on gluconeogenesis was concentration-dependent since the infusion of 2 mM NH&l together with octanoate did not overcome the inhibition caused by the fatty acid until NH&l infusion was terminated. Mitochondria were incubated as described in Fig. 1 except that 2 mM malate and 2 rnM a-ketoglutarate were added together to each incubation, NH,CI, when present, was added at a concentration of 2 mM, cycloheximide at 5 mM, octanoate at 0.2 mM, and L-palmitylcarnitine at 0.1 mM. Values are the means rt standard error for three experiments. a shift in the mitochondrial oxidation-reduction state toward oxidation as reflected by a change in the acetoacetate to fihydroxybutyrate ratio from 4 before, to approximately 6, after 15 min NH,Cl infusion. Urea synthesis was sharply increased by NH,Cl infusion in a manner roughly proportional to the concentration of NH&l. Thus 0.9 rmol of urea/min/g of liver were formed during infusion of 1 mM NH&l while 1.5 pmol were noted with 2 mM NH&l.
In order to clarify the differential effect of 1 and 2 mM NH&l on gluconeogenesis, livers were "freeze-clamped" before, during, and after NH,Cl infusion and various intermediates were measured (Table II). Two amino acids, aspartate and alanine, both in relatively low concentration in the liver, were increased 5-to &fold after the infusion of 1 mM NH&l and 20-to 30-fold when 2 mM NH&l was infused for 15 min. The glutamate concentration in the perfused liver was approximately doubled by 2 mM NH&l but unaffected at the lower concentration.
The concentrations of the adenine nucleotides were not significantly altered at either 1 or 2 mM NH&l. An analysis of glycine (Table  II) and of all of the other amino acids (data not shown) in the liver indicated no major changes associated with NH&I infusion. The high concentrations of both aspartate and alanine in perfused guinea pig liver dropped markedly immediately after cessation of 2 mM NH&l infusion. This decrease corresponded to a marked increase in gluconeogenesis noted after the termination of NH&l (2 mM) shown in Fig. 6, suggesting a direct relationship between the two events. There was a net loss of 6.2 rmollg of liver from the three major amino acids (glutamate, aspartate, and alanine) observed to change in the 15-min period after cessation of 2 mM NH,Cl infusion. Negligible amounts of amino acids were released by the liver into the perfusion medium and the slight increase in oxygen consumption noted during this period was not sufficient to account for their subsequent oxidation in the citric acid cycle. This net loss in amino acids could account for 3.1 Mmol of glucose and 3.1 pmol of urea/g of liver in 15 min if converted entirely to these two major end products by the liver. A calculation of the rate of glucose synthesis during the 15-min period immediately after the cessation of 2 mM NH&l infusion indicated that approximately 3 rmol of glucose/g of liver were formed over the depressed level of gluconeogenesis caused by perfusion with octanoate alone. Urea synthesis declined more rapidly than glucose synthesis increased but net urea synthesis during the same 15-min period was approximately 3 pmol/g of liver. The rapid changes in the rates of urea and glucose synthesis noted immediately after removal of NH,Cl complicate the exact calculation of carbon and nitrogen balance but it is probable that the majority of the amino acids stored in the liver are converted to glucose, accounting for the sharp increase in gluconeogenesis.
Further information about the pathway of carbon flow during gluconeogenesis in guinea pig liver is presented in Fig. 7. Octanoate infusion caused the expected decrease in glucose synthesis of about 50% and the combination of 0.2 mM aminoxyacetate, and 2 mM NH&l reduced gluconeogenesis to values noted prior to infusion of lactate. Urea synthesis increased after NH&l infusion and returned to control levels within 5 min of the termination of NH&l infusion. However, the rate of gluconeogenesis did not rebound to the high levels noted after NH&l infusion in the absence of aminoxyacetate (see Fig.  6). After 95 min of perfusion, octanoate infusion was terminated and the synthesis of glucose immediately returned to levels approximately equal to those noted before the introduction of aminoxyacetate and NH&l. This increase in glucose synthesis reflected the formation of P-enolpyruvate by mitochondrial Penolpyruvate carboxykinase, a pathway not inhibited by aminoxyacetate (8). An analysis of the concentration of the two amino acids which changed dramatically during NH&l in- Livers from 48-hour fasted guinea pigs were perfused as described in Fig. 6. 8983 fusion, glutamate and aspartate, indicated that glutamate was present in guinea pig liver at a level of 5.20 pmol/g of liver during the infusion with aminoxyacetate and NH&I and did not change when the infusion of these two compounds was terminated.
Aspartate, on the other hand, dropped from a concentration of 1.35 Kmol/g of liver to 0.29 pmol after the infusion of aminoxyacetate and NH&l. It is probably that aminoxyacetate, by inhibiting aspartate aminotransferase prevents the buildup of aspartate in the liver. This blocks the normal rebound in gluconeogenesis shown in Fig. 6 after cessation of infusion of 2 mM NH&I.
The effect of aminoxyacetate did not appear to be complete, however, since the concentration of aspartate in the liver rose from a basal value of 0.153 pmol/g of liver prior to NH&l infusion to 1.35 rmol despite the presence of the inhibitor. This increase in aspartate may support the urea cycle which is functioning at an elevated rate despite the inhibition of gluconeogenesis by aminoxyacetate.

DISCUSSION
It is well documented that the oxidation of fatty acids by guinea pig liver inhibits gluconeogenesis from a number of substrates. Studies with isolated guinea pig liver mitochondria have demonstrated a close correlation between the control of P-enolpyruvate formation and control of gluconeogenesis in the intact, perfused liver (4,7,8,32). Fatty acid oxidation by isolated guinea pig liver mitochondria markedly diminishes P-enolpyruvate formation, an effect shown to be directly related to the oxidation-reduction state of the mitochondria (Table I)
When aspartate aminotransferase was blocked by addition of aminoxyacetic acid, there was a 4-fold stimulation of P-enolpyruvate synthesis from cu-ketoglutarate which was markedly diminished by octanoate, oleate, or octanoylcarnitine.
These findings offer further support for the role of the NAD+/NADH ratio in modulating the rate of mitochondrial P-enolpyruvate formation. In this study we offer evidence that the effect of fatty acids on P-enolpyruvate synthesis from either a-ketoglutarate or malate can be reversed by the addition of NH&l to the mitochondrial incubation medium. This is presumably due to a displacement of the glutamate dehydrogenase equilibrium toward glutamate synthesis, resulting in an increase in the NAD+/NADH ratio. This is consistent with an increased rate of glutamate formation and release by the mitochondria. Aspartate also increases due to transamination with oxalacetate generated from the substrate malate. It is of interest that the addition of aminoxyacetate acid to the incubation medium together with NH&l (in the absence of exogenous fatty acids) resulted in the highest rate of P-enolpyruvate formation noted in this study, but at the same time markedly reduced the level of transamination of oxalacetate to aspartate. It is apparent that, if the level of glutamate is increased by the addition of NH&l, aspartate aminotransferase can effectively compete with P-enolpyruvate carboxykinase and divert oxalacetate toward aspartate formation.
The rate of transamination of oxalacetate is remarkably constant (see Table I) even when glutamate is increased to very high values by the addition of cycloheximide, an inhibitor of energy transfer at site 1. The use of cycloheximide with isolated guinea pig liver mitochondria offers further insight into the relationship between the oxidation-reduction state and P-enolpyruvate synthesis. Previous studies by  demonstrated a virtually complete inhibition of hepatic gluconeogenesis by cycloheximide infusion into isolated guinea pig liver. Cycloheximide apparently causes this effect by interacting at site 1 of the respiratory chain, causing an increase in the level of hepatic NADH and a decrease in the concentration of ATP. Cycloheximide added together with octanoate or L-palmitylcarnitine markedly reduces the rate of P-enolpyruvate formation and lowers ATP synthesis by isolated guinea pig liver mitochondria.
Interestingly, the rate of total ketone body formation from octanoate was stimulated 2-fold by cycloheximide addition, whereas the ratio of acetoacetate to &hydroxybutyrate by isolated guinea pig liver mitochondria was decreased. Lopes-Cardozo and Van Den Bergh (33) have shown an inverse relationship between the concentration of citric acid cycle intermediates, such as malate, and the synthesis of ketone bodies from palmitate by rat liver mitochondria. They also noted a decrease in the acetoacetate to &hydroxybutyrate ratio when the concentration of citric acid cycle intermediates was increased. There have been suggestions in the literature that fatty acids stimulate P-enolpyruvate synthesis in rabbit or guinea pig liver by uncoupling mitochondrial respiration (34-36). In the present study we show no uncoupling of guinea pig liver mitochondria by octanoate at concentrations of 0.5 mM or below. At a higher concentration of octanoate, 1 mM, the mitochondria are loosely coupled. It is known that 2,4-dinitrophenol will stimulate P-enolpyruvate formation from a number of citric acid cycle intermediates but in those experiments ATP was added to the incubation medium. While it is possible that uncoupling of oxidative phosphorylation by fatty acids plays a physiological role in the regulation of hepatic gluconeogenesis, it is an unlikely explanation for the marked decrease in glucose synthesis noted in these papers. Fatty acids are also known to inhibit ATP-ADP exchange as well as alter the kinetics of tricarboxylate carriers (37-41). It is also possible that fatty acids, such as octanoate, might reduce the rate of P-enolpyruvate formation and subsequent efflux from the mitochondria by such a mechanism. However, palmitylcarnitine, which is one of the least effective fatty acyl derivatives in inhibiting the various tricarboxylate carriers, very effectively decreases P-enolpyruvate synthesis. Our data are, therefore, consistent with Morel et al. (42)  is elevated artificially, such as by the perfusion of 2 mM NH&l, it is clear that some of the glutamate may be deaminated, supporting urea synthesis. There was a net decrease of 3.3 rmol of glutamate within the 15 min after termination of NH&l (Table II) which could not be accounted for by release into the perfusion medium or by transamination to form other amino acids.
It is possible that during fasting the oxidation-reduction state of guinea pig liver mitochondria can be influenced by the metabolism of ammonia as well as fatty acids. Our liver perfusion experiments using ammonium chloride together with fatty acids demonstrate a potential interaction which can dramatically alter the pattern of regulation of glucose synthesis in the liver. This observation also provides a link between two previously unresolved observations: the oxidized mitochondrial NAD+/NADH ratio during fasting and the high rates of fatty acid oxidation which normally characterize starvation.
It is probable that the availability of ATP will limit the rate of disposal of ammonia through the urea cycle. Our studies (Fig. 6) indicate a lack of exact stoichiometry between the rate of urea synthesis and the concentration of NH&l in the perfusion medium. Urea synthesis from 2 mM NH&l is not double that found with 1 mM NH&l, suggesting some limitation on this process. However, there is no relationship between the rate of gluconeogenesis and ureogenesis. With 2 mM NH,Cl infusion, glucose synthesis remains at levels noted prior to its infusion and does not increase until NH&l infusion is terminated. One important aspect of the interaction between ammonia and fatty acid metabolism is the key role of aspartate and, to a lesser extent, alanine as potential reservoirs of carbon for glucose synthesis. When levels of ammonia high enough to temporarily exceed the capacity of the urea cycle were used (2 mM NH&l), the concentrations of aspartate and alanine increased 20-and 30-fold, respectively.
The total aspartate concentration was a remarkably high 6 mM (assuming the intracellular water concentration is approximately 50% of total (45)). The rapid drop in the aspartate levels after cessation of NH&l infusion into guinea pig liver, which correlates with an increase in gluconeogenesis, suggests that the aspartate carbon is converted to glucose. The negligible amounts of aspartate detectable in the perfusate indicate that this decrease in cellular aspartate is not due to release of the amino acid from the cell. These findings are in general accord with a *Unpublished observations.