Main

JVS mice were first reported in a C3H.OH strain (formerly named C3H-H-2°)(1). These mice are autosomal recessive mutants that exhibit hypoglycemia, hyperammonemia, and severe fatty liver in the infantile period. Growth retardation appears 3 wk after birth(2), and cardiac hypertrophy is also found at the age of 2 to 3 mo(3). Kuwajima et al.(4) discovered that these mice have severe systemic carnitine deficiency, which is caused by an abnormality of the renal carnitine transport system(5) that is also seen in human systemic carnitine deficiency. Further, carnitine administration to these mice corrects both the growth retardation and the cardiac hypertrophy(3, 6) as in human systemic carnitine deficiency.

The major role of carnitine is the transport of long chain acyl-CoA compounds across the mitochondrial inner membrane(7, 8). Long chain acyl-CoA compounds penetrate the mitochondrial outer membrane and are converted into long chain acylcarnitine by the action of CPT I. They move through the mitochondrial inner membrane by carnitine-acylcarnitine translocase and are reconverted into acyl-CoA compounds by CPT II in the mitochondrial matrix(8, 9). Carnitine also participates in modulation of the intramitochondrial acyl-CoA/CoASH ratio by pooling and transporting short chain acylesters(10). Carnitine deficiency in humans is classified into two categories, primary and secondary carnitine deficiency(11). Secondary carnitine deficiency is caused by various pathophysiologic conditions, such as mitochondrial β-oxidation disorders or the administration of valproic acid(12) and some antibiotics(13). Premature infants also have reduced or borderline serum carnitine levels(14, 15). Carnitine supplement therapy is occasionally used, but whether the therapy is necessary is unclear(16). Primary carnitine deficiency is defined as a genetic defect in the renal reabsorption of carnitine(11). Most cases of primary carnitine deficiency in humans exhibit low serum ketone bodies, but unchanged or increased serum ketone bodies have also been reported(11). The question arises as to how much carnitine in tissue or serum is needed for β-oxidation and ketogenesis.

Because JVS mice are thought to be a true murine counterpart of human systemic carnitine deficiency, they are useful animal models for studying the physiologic role of carnitine. In the present study, we examined the effects of carnitine and fatty acids of various chain lengths on ketogenesis and the relationship between ketogenesis and liver carnitine concentration using the liver perfusion technique.

METHODS

Animals. Control mice (normal homozygotes) and JVS mice(homozygous mutants) were identified as described previously(3, 5). JVS mice were injected intraperitoneally with 5 μmol of L-carnitine daily from 11 to 20 d after birth to increase the survival rate. Control and JVS mice, varying from 9 to 13 wk after birth, were starved for 24 h before the perfusion experiments, and body weights were measured.

Technique of liver perfusion. After anesthesia with an intraperitoneal injection of sodium pentobarbital (0.05 mg/g, body weight), livers were perfused in situ with Krebs-Henseleit bicarbonate buffer(pH 7.4, 36 °C, saturated with 95% oxygen and 5% carbon dioxide) in a nonrecirculating system(17). A Clark-type oxygen electrode (INSTECH, model 125/05) was used to measure oxygen concentration of effluent perfusate. Perfusate and substrate were pumped into the liver through a cannula inserted in the portal vein, and effluent perfusate was collected via a cannula placed in the inferior vena cava(17, 18). The livers were perfused for 10 min at a constant flow rate of about 8 mL/min. Perfusion was considered successful when the liver did not swell or leak perfusate and oxygen consumption was stable. Oleate, a long chain fatty acid, octanoate, a medium chain fatty acid, and butyrate, a short chain fatty acid, were all dissolved or bound in a 20% BSA solution and were infused into the liver as substrates diluted to a final concentration of 0.3 mM 0.33% BSA solution. Previously, we found that 0.3 mM oleate gives about the maximum rate of ketone body production in normal perfused liver (data not shown). L-Carnitine was dissolved in Krebs-Henseleit buffer at final concentrations of 50, 100, 200 and 1000 μM and infused into the liver.

Analysis of liver water. The determination of total, extracellular, and intracellular water in the liver was carried out according to a method described previously(19) with the following modifications. Livers from control and JVS mice were perfused with[14C]inulin and [3H]water for 5 min. Extracellular water in the perfused liver was measured by determining the distribution of[14C]inulin in the perfusion medium and a portion of homogenized perfused liver. Total water was measured by drying another portion of the liver at 50 °C for 7 d. The data acquired by this method were in close accordance with data using the distribution of [3H]water in the medium and the perfused liver (data not shown). Intracellular water was then calculated by subtracting extracellular from total water.

Liver carnitine concentration. At the end of perfusion experiments, each liver was rapidly weighed and pressed between metal clamps previously cooled in liquid nitrogen, and then stored at -60 °C. For free carnitine determination, the liver was homogenized and deproteinized with 60% perchloric acid. For total carnitine determination, carnitine esters in the supernatants of the homogenates were hydrolyzed by 10 N KOH and then deproteinized with perchloric acid. Levels of total and free carnitine in neutralized supernatants were measured using an enzymatic cycling technique with carnitine dehydrogenase(20).

Analysis of perfusion medium. Samples (3 mL) from effluent perfusate were taken at 2-min intervals during the experiment and were immediately deproteinized with 0.158 mL of 60% perchloric acid. After centrifugation and neutralization, 3-hydroxybutyrate, and acetoacetate were measured enzymatically. Total ketone body was defined as the sum of 3-hydroxybutyrate and acetoacetate, and the rate of ketone body production was calculated according to the formula: Rate of ketone body production(μmol/h/whole liver) = total ketone body (μmol/mL/whole liver) × flow rate (mL/h) The rate per whole liver was used as a unit in the experiments.

Statistical analysis. The significance of the differences between means was established by an unpaired t test.

RESULTS

Body weight and liver weight. Seven control mice and 25 JVS mice were studied. Although the body weights in JVS mice were similar to those in control mice, the livers from JVS mice were about twice as heavy. The ratio of liver weight to body weight also increased in JVS mice compared with control mice (Table 1).

Table 1 Body weight and liver weight in control and JVS mice
Full size table

Amount of liver water. Table 2 shows the total, extracellular, and intracellular water in livers from control and JVS mice. The total water in livers from JVS mice was much lower than that from control mice (444 ± 37 versus 708 ± 26 μL/g of wet liver), whereas the extracellular water in JVS mice was not significantly different from that in control mice. The intracellular water in livers from JVS mice was therefore lower than that from control mice (292 ± 49versus 579 ± 47 μL/g of wet liver).

Table 2 Total, extracellular, and intracellular water in livers from control and JVS mice
Full size table

Carnitine concentration in livers. Table 3 shows the concentrations of carnitine in livers from control and JVS mice after 10 min of perfusion in the absence of added substrates. The value of carnitine concentration in effluent perfusate was lower than the detection limit (less than 0.2 nmol/mL). Total and free carnitine concentrations in the livers of JVS mice were only 3 and 4%, respectively, of those of control mice. The ratio of acylcarnitine to free carnitine was also lower in JVS mice than in control mice.

Table 3 Carnitine concentration in perfused livers from control and JVS mice
Full size table

Effect of carnitine on ketogenesis from oleate, octanoate, and butyrate. Figure 1 shows the change of the rate of ketone body production after infusion of 0.3 mM oleate, octanoate, or butyrate and 1000 μM L-carnitine into livers from control and JVS mice. In control mice, infusion of each of the three different chain length fatty acids rapidly increased the rate of ketone body production about 2-fold, and it decreased after infusion ended. The infusion of L-carnitine did not have any effect on ketone body production (Fig. 1, A-C). In JVS mice, the infusion of oleate did not increase ketone body production, but the additional infusion of L-carnitine increased production about 2.5 times. After infusion of L-carnitine into livers of JVS mice for 10 min, the rate of ketone body production remained at the same level for at least 16 min after the infusion ended (Fig. 1A). The infusion of octanoate or butyrate increased ketone body production in JVS mice, and the additional infusion of L-carnitine caused a further increase in production (Fig. 1, B and C).

Figure 1
figure 1

Effect of carnitine on ketogenesis in perfused livers of JVS and control mice after infusion of fatty acids. Livers from fasted JVS(•: n = 3) and control (○: n = 3) mice were perfused as described under “Methods.” After perfusion with buffer alone for ten minutes, 0.3 mM oleate (A), octanoate (B), or butyrate (C) was infused into the livers for 30 min. Ten minutes after the infusion of each fatty acid began, 1000 μM L-carnitine was infused for 10 min. Values are means ± SE.

Full size image

Effect of oleate, octanoate, or butyrate on ketogenesis in the presence of carnitine. As shown in Figure 2, 1000μM L-carnitine was first infused into the liver, followed by infusion of 0.3 mM oleate, octanoate, or butyrate. In control mice, the infusion of L-carnitine did not increase the rate of ketone body production, but the subsequent infusion of each fatty acid rapidly increased it about 2-fold (Fig. 2, A-C). In JVS mice, 1000 μM L-carnitine rapidly increased the rate of ketone body production about 2.5-fold, and it remained at the same level for 10 min. The subsequent infusion of oleate did not increase the rate of ketone body production (Fig. 2A), whereas octanoate and butyrate resulted in a further increase (Fig. 2, B and C).

Figure 2
figure 2

Effect of oleate, octanoate, and butyrate infusion on ketogenesis in perfused livers of JVS and control mice in the presence of carnitine. Livers from fasted JVS (•: n = 3) and control(○: n = 3) mice were perfused as described under“Methods.” After perfusion with buffer alone for 10 min, 1000μM L-carnitine was infused into the livers for 30 min. Ten minutes after L-carnitine infusion was initiated, 0.3 mM oleate (A), octanoate(B), or butyrate (C) was infused for 10 min. Values are expressed as means ± SE.

Full size image

Comparison of the rate of ketone body production. The rates of ketone body production per whole liver from control mice (n = 11) and JVS mice (n = 31) before the infusion of fatty acids or L-carnitine were 40.1 ± 14.7 and 49.0 ± 12.6 μmol/h/whole liver, respectively, and did not differ significantly. The rate of ketone body production in JVS mice after the infusion of L-carnitine was significantly higher than in control mice after the infusion of oleate (130.4 ± 19.4(n = 6) versus 79.6 ± 16.1 (n = 6)μmol/h/whole liver) (Figs. 1A and 2A).

Relationship between carnitine concentration in perfusate and ketogenesis in livers from JVS mice. Figure 3 shows the results of liver perfusion with different concentrations of L-carnitine for 10 min in the absence of added fatty acids. The infusion of 1000 μM L-carnitine rapidly increased the rate of ketone body production, and the infusion of lower concentrations (200, 100, and 50 μM) of L-carnitine increased the rate of ketone body production in a dose-dependent manner. Ketone body production remained unchanged for at least 6 min after the infusion of L-carnitine ended.

Figure 3
figure 3

Effect of various levels of infused carnitine on ketogenesis in JVS mice. L-Carnitine, 50 μM (n = 3), 100 μM(n = 3), 200 μM (n = 3), or 1000 μM (n = 4), was infused into livers from JVS mice for 10 min. Fatty acids were not infused in this experiment. Values are expressed as means ± SE.

Full size image

Relationship between free carnitine concentration in liver and ketogenesis in JVS mice. Figure 4A shows the relationship between the concentration of free carnitine in the liver and the rate of ketone body production in JVS mice. The horizontal axis shows the free carnitine concentration in liver from JVS mice. The carnitine concentration in the effluent perfusate decreased rapidly after stopping the infusion of L-carnitine and became below the detection limit in 2 min. The vertical axis is the rate of ketone body production upon stopping the infusion of various concentrations of L-carnitine. The rate of ketone body production was low when L-carnitine was not infused, and increased proportionally to the increase of free carnitine concentration in the liver due to the infusion of L-carnitine. However, when the free carnitine concentration exceeded about 100 nmol/g of wet liver, the rate of ketone body production ceased to increase and remained at about the same level. The values of the maximal rate of ketone body production (Vmax) and the value of carnitine concentration required for half-maximal ketogenesis (Km for free carnitine) were calculated to be 152 μmol/h/whole liver and 13 nmol/g of wet liver, respectively (Fig. 4B). The carnitine concentrations required for 80 and 90% Vmax were also calculated as 52 and 120 nmol/g of wet liver, respectively. Based on the amount of intracellular water in the liver from JVS mice(Table 2), 50% Vmax(Km for free carnitine), 80% Vmax, and 90% Vmax were calculated as 45, 178, and 411 μM, respectively.

Figure 4
figure 4

Relationship between free carnitine concentration in liver of JVS mice and ketogenesis. (A) Free carnitine concentrations in livers from JVS mice were measured at the end of the perfusion experiment as described under “Methods.” Livers were perfused with perfusate alone for 6 min to wash out L-carnitine in the extracellular space. The rate of ketone body production was assessed when the infusion of various concentrations of L-carnitine was stopped. (B) Lineweaver-Burk plot assuming that A shows an enzyme reaction in which free carnitine is a substrate and ketone body is a product in livers from JVS mice.

Full size image

DISCUSSION

JVS mice were recently discovered(1, 2) and are characterized by carnitine deficiency(4). The value of total carnitine concentrations in the liver of JVS mice was 11.6 ± 2.6 nmol/g of wet liver, which was only 3% of that of control mice. The value obtained from the present study was lower than that previously reported by Kuwajima et al.(4) (37.4 ± 5.9 nmol/g of wet liver, 4 wk after birth). The difference may be attributed to age after birth, duration of fasting, carnitine assay method, or carnitine release into perfusate during perfusion(21).

We examined the concentration of carnitine required for ketogenesis using fatty acids with different chain lengths in perfused livers from control and JVS mice. Experiments using liver perfusion are advantageous in that they are more similar to in vivo conditions than those using isolated mitochondria or hepatocytes, and it is possible to add substrates accurately and to measure products in the effluent perfusate directly.

Effect of carnitine and fatty acids on ketogenesis. Ketone body production in the liver of JVS mice was not affected at all by the infusion of a long chain fatty acid alone, and the subsequent infusion of carnitine increased ketogenesis (Fig. 1A). Furthermore, the infusion of carnitine alone caused an increase in ketogenesis which was not affected by the subsequent infusion of a long chain fatty acid (Fig 2A). These results suggested that the substrate of ketone body production of JVS mice in the presence of carnitine is endogenous long chain fatty acids, and indicated that the rate-limiting factor for ketogenesis in the liver of JVS mice is carnitine, not fatty acids. On the other hand, ketogenesis in the liver of control mice is limited by the supply of fatty acids, but not by carnitine, as shown in Figs. 1A and 2A. The infusion of short or medium chain fatty acid caused a stimulation of ketogenesis in JVS mice without carnitine infusion (Fig. 1, B and C), clearly indicating that ketogenesis from these medium and short chain fatty acids does not require carnitine. This result agrees with previous publications(22). It is reasonable to consider that the further increase of ketogenesis by the infusion of carnitine during the infusion of short or medium chain fatty acids is caused by the carnitine-stimulated metabolism of endogenous long chain fatty acids. The results shown in Fig. 2, B and C, are consistent with this conclusion.

Kuwajima et al.(4) reported that total lipid in the liver of JVS mice reached about 44% of total liver weight compared with 6% in control mice, and that only the triglyceride level in liver of JVS mice was about seven times higher than in control mice. Therefore, we also suggest that the greater liver weight of JVS mice, which is about twice that of control mice, results from the accumulated triglycerides, and that endogenous long chain fatty acids from the triglycerides are used as substrates in the presence of carnitine in JVS mice. In this study, the rate per whole liver was used as a unit of the rate of ketone body production, not the rate per g of liver. That is because livers of JVS mice were about twice as heavy as those of control mice, mainly due to lipid accumulation, although the body weight of JVS and control mice was almost the same.

There was no significant difference in the basal rate of ketone body production between JVS and control mice. The level of the basal rate of ketogenesis in JVS mice may be indispensable for survival; in fact, JVS mice often die during fasting. On the other hand, the value of maximal ketogenesis in the presence of carnitine was significantly higher in JVS mice than control mice. This result is consistent with the previous report by Imamura et al.(23) showing that the rate of β-oxidation from palmitate with carnitine in liver homogenate is much higher in JVS mice than control mice. Impairment of the transport of long chain acyl-CoAs into mitochondria in JVS mice may result in increased activity of the carnitine-dependent transport system and mitochondrial β-oxidation enzymes.

Carnitine level in liver and ketogenesis. The rate of ketogenesis in JVS mice increased in a dose-dependent manner just after the infusion of L-carnitine and showed saturation kinetics (Figs. 3 and 4A). Infused L-carnitine entered liver cells, and the rate of ketone body production increased according to the free carnitine concentration in the cells. The free carnitine concentrations in liver cells required for 80 and 90% of maximal ketogenesis of JVS mice were 178 and 411 μM, respectively. The free carnitine concentration in liver cells of control mice (357 μM; Table 3) gave 89% of the maximal value of ketogenesis, and ketone body production was saturated (Fig. 4A). The present results show that ketone body production from long chain fatty acids is almost unchanged, even when the free carnitine concentration is decreased to 50% of the normal level. On the other hand, a small increase of carnitine has a great effect on ketone body production in carnitine-deficient JVS mice, because ketogenesis is not saturated for carnitine in JVS mice. In addition, JVS mice have a higher maximal rate of ketone body production than do control mice, which may be caused by increased intramitochondrial β-oxidation activity (Figs. 1 and 2). The maximum rate of ketone body production after the addition of oleate in control mice (79.6 μmol/h/whole liver) could be calculated to the free carnitine concentration of about 50μM in the liver cells of JVS mice, which was only 14% of the value in control mice.

Many enzymes participate in the β-oxidation pathway from the transport of acyl-CoA compounds into mitochondria to the respiratory chain or ketone body production. Carnitine-dependent enzymes involved in the transport of long chain acyl-CoA compounds into mitochondria are CPT I, CPT II, and carnitine-acylcarnitine translocase. McGarry et al.(24, 25) proposed that CPT I could be the primary control site in the physiologic regulation of hepatic mitochondrialβ-oxidation. Brady et al.(26) reported that there was a close relationship between CPT I activity andβ-oxidation. McGarry et al.(27) also reported that Km values of CPT I from rat and human fetal liver for L-carnitine were 32 and 39 μM, respectively. These values are close to the apparent Km value from several reactions in the perfused livers of the present study. This suggests that the value of free carnitine required for half-maximal ketogenesis in the perfused liver from JVS mice corresponds to the Km value of CPT I for carnitine and that the availability of carnitine limits fatty acid oxidation in systemic carnitine deficiency.

We conclude that carnitine is a rate-limiting factor for β-oxidation in liver only when the carnitine level in liver is very low and that the perfusion technique applied to the liver of JVS mice is useful for understanding the physiologic and pathologic roles of carnitine.