Use of control analysis to study the regulation of plant lipid biosynthesis

Control analysis is a powerful method to quantify the regulation of metabolic pathways. We have applied it to lipid biosynthesis for the first time by using model tissue culture systems from the important oil crops, olive ( Olea europaea L.) and oil palm ( Elaeis guineensis Jacq.). By the use of top-down control analysis, fatty acid biosynthesis has been shown to exert more control than lipid assembly under different experimental conditions. However, both parts of the lipid biosynthetic pathway are important, so that attempts to alter oil yield by manipulating the activity of a single enzyme step are very unlikely to produce significant increases.

Carnitine palmitoyl transferase I and the control of myocardial beta-oxidation flux

Carnitine palmitoyltransferase I is assumed to be rate limiting for beta-oxidation in all tissues. However, the concentration of malonyl-CoA in heart and muscle is high and is enough to completely inhibit beta-oxidation if this assumption is correct. In this review, we consider whether: (i) there is a malonyl-CoA-insensitive carnitine palmitoyltransferase I activity; (ii) the measured malonyl-CoA concentration in the heart is physiologically meaningful; and (iii) carnitine palmitoyltransferase I is rate-limiting for beta-oxidation in the heart.

Carnitine palmitoyl transferase I and the control of beta-oxidation in heart mitochondria

Mitochondrial beta-oxidation provides much of the fuel requirements of heart and skeletal muscle despite the malonyl-CoA concentration greatly exceeding the IC(50) of carnitine palmitoyl transferase for malonyl-CoA. To try to explore the relationship between inhibition of carnitine palmitoyl transferase I activity and beta-oxidation flux, we measured the flux control coefficient of carnitine palmitoyl transferase I over beta-oxidation carbon flux in suckling rat heart mitochondria. The flux control coefficient was found to be 0.08 +/- 0.05 and 50% of carnitine palmitoyl transferase I activity could be inhibited before beta-oxidation flux was affected. These observations may help to explain the presence of high rates of beta-oxidation despite the high concentration of malonyl-CoA in rat heart; we hypothesize that although not rate-limiting in vitro, carnitine palmitoyl transferase is rate-limiting in vivo because of the high malonyl-CoA concentration in heart and muscle.

Effect of lipopolysaccharide and cytokines on oxidative metabolism in neonatal rat hepatocytes

BACKGROUND/PURPOSE

Lipopolysaccharide (LPS) and cytokines produced during neonatal sepsis trigger free radical production, which eventually results in inhibition of liver metabolism. Studies in adults have indicated a hypermetabolic response to sepsis; however, evidence for a hypermetabolic response in neonates is equivocal. This study was carried out to determine whether LPS and cytokines can cause liver hypermetabolism in neonates.

METHODS

The initial bacterial insult and cytokine cascade were mimicked by the addition of lipopolysaccharide (Escherichia coli 055:B5), tumour necrosis factor (TNF-alpha), and interleukin-6 (IL6) during the isolation of hepatocytes by collagenase digestion from 11- to 13-day-old Wistar rats. Hepatocyte oxygen consumption was measured polarographically with cells respiring on palmitate (0.5 mmol/L). Myxothiazol, a specific inhibitor of mitochondrial respiration, was used to distinguish extra- and intramitochondrial oxygen consumption. Morphologic changes were assessed by electron microscopy.

RESULTS

The addition of LPS, TNF-alpha and IL6 during hepatocyte isolation resulted in a 10% decrease in cell yield (P <.05) compared with untreated controls; however, cell viability was unchanged (n = 31). Both total and extramitochondrial oxygen consumption were significantly greater in treated cells compared with untreated controls (P <.05, Student's t test). Electron microscopy indicated that LPS, TNF-alpha, and IL6 did not cause ultrastructural changes to hepatocytes.

CONCLUSIONS

The increase in oxygen consumption was predominantly extramitochondrial and likely to be caused by increased oxygen requirement for cytosolic detoxification and repair purposes. This study shows that liver hypermetabolism metabolism can occur in response to LPS and cytokines. However, during in vivo neonatal sepsis, additional free radical damage may blunt this hypermetabolic response.

Neonatal oxidative liver metabolism: effects of hydrogen peroxide, a putative mediator of septic damage

BACKGROUND/PURPOSE

Surgical neonates are at risk for sepsis and liver dysfunction. These complications are more common in preterm neonates and in those who receive total parenteral nutrition. Elevated levels of reactive oxygen species (eg, hydrogen peroxide) have been reported in these "at-risk" patients and may be the mediators of liver impairment via their effect on oxidative energy metabolism. The aim of this study was to test the hypothesis that elevated levels of hydrogen peroxide (H2O2) impair neonatal liver oxidative energy metabolism.

METHODS

An in vitro model to test this hypothesis was developed in hepatocytes isolated from neonatal (11-day to 15-day) rats. The cells, respiring on palmitate (0.5 mmol/L in 2% bovine serum albumin), were exposed to H2O2. Oxygen consumption was measured polarographically. In experiment A, H2O2 was added to the cell preparation at different concentrations (0.5 mmol/L, 1 mmol/L, 1.5 mmol/L, 2 mmol/L) to assess the effect on oxygen consumption. In experiment B, H2O2 (2 mmol/L) was added to hepatocytes in the presence of inhibitors of mitochondrial respiration to define the site of action of H2O2. In experiment C, electron microscopy was performed on hepatocytes after incubation with 1 mmol/L and 2 mmol/L of H2O2.

RESULTS

In experiment A, H2O2 significantly reduced hepatocyte oxygen consumption at 1.5 and 2 mmol/L. In experiment B, in the presence of inhibitors of mitochondrial respiration, myxothiazol (inhibitor of substrate oxidation), and oligomycin (inhibitor of adenosine triphosphate (ATP) synthase), no further inhibition by H2O2 occurred, indicating that the effect of H2O2 was intramitochondrial and affecting the synthesis of ATP. In experiment C, microscopic alterations of mitochondria were noticed exclusively in hepatocytes incubated with 2 mmol/L H2O2.

CONCLUSIONS

Results of this study demonstrate that H2O2 impairs neonatal liver oxidative metabolism. H2O2 probably directly inhibits ATP synthase. The authors hypothesize that H2O2 may play a role in the biochemical pathogenesis of liver dysfunction associated with sepsis. Identification of the precise target site of H2O2 may be valuable in directing therapy in septic neonates.

Analgesic doses of fentanyl impair oxidative metabolism of neonatal hepatocytes

BACKGROUND/PURPOSE

Studies in human surgical neonates have shown that intraoperative fentanyl analgesia results in greater fall in perioperative body core temperature compared with morphine analgesia. The aim of the study was to compare in a neonatal animal model the biochemical effect of fentanyl and morphine on hepatocyte oxidative metabolism.

METHODS

Hepatocytes were isolated from suckling rats and the oxygen consumption from palmitate was measured polarographically. In experiment A, fentanyl and morphine within the respective analgesic serum ranges were added to hepatocytes to assess the effect on oxygen consumption. In experiment B, fentanyl was added to hepatocytes in the presence of inhibitors of mitochondrial respiration to investigate its site of action. In experiment C, hepatocytes were incubated with either fentanyl or morphine, centrifuged, and then examined ultrastructurally by electron microscopy.

RESULTS

In experiment A, fentanyl inhibited oxygen consumption by up to 40% (P < .01). Morphine inhibited oxygen consumption to a maximum of 25% (P < .01). In experiment B, in the presence of oligomycin, fentanyl increased the inhibition of oxygen consumption; however, in the presence of myxothiazol, no further inhibition by fentanyl occurred. In experiment C, mild ultrastructural alterations to hepatocytes were observed after incubation with fentanyl but not with morphine.

CONCLUSIONS

This study demonstrates that therapeutic doses of two commonly used analgesic drugs impair neonatal hepatic oxidative metabolism. Fentanyl exerts a greater effect than morphine by diminishing liver oxygen consumption by up to 40%. The inhibitory effect of fentanyl occurs directly on the mitochondrial respiratory chain, either on substrate oxidation or on the thermogenic proton leak. The findings of this study are relevant to the perioperative management of surgical neonates.

Comparisons of flux control exerted by mitochondrial outer-membrane carnitine palmitoyltransferase over ketogenesis in hepatocytes and mitochondria isolated from suckling or adult rats

The primary aim of this paper was to calculate and report flux control coefficients for mitochondrial outer-membrane carnitine palmitoyltransferase (CPT I) over hepatic ketogenesis because its role in controlling this pathway during the neonatal period is of academic importance and immediate clinical relevance. Using hepatocytes isolated from suckling rats as our model system, we measured CPT I activity and carbon flux from palmitate to ketone bodies and to CO2 in the absence and presence of a range of concentrations of etomoxir. (This is converted in situ to etomoxir-CoA which is a specific inhibitor of the enzyme.) From these data we calculated the individual flux control coefficients for CPT I over ketogenesis, CO2 production and total carbon flux (0.51 +/- 0.03; -1.30 +/- 0.26; 0.55 +/- 0.07, respectively) and compared them with equivalent coefficients calculated by similar analyses [Drynan, L., Quant, P.A. & Zammit, V.A. (1996) Biochem. J. 317, 791-795] in hepatocytes isolated from adult rats (0.85 +/- 0.20; 0.23 +/- 0.06; 1.06 +/- 0.29). CPT I exerts significantly less control over ketogenesis in hepatocytes isolated from suckling rats than those from adult rats. In the suckling systems the flux control coefficients for CPT I over ketogenesis specifically and over total carbon flux (< 0.6) are not consistent with the enzyme being rate-limiting. Broadly similar results were obtained and conclusions drawn by reanalysis of previous data {from experiments in mitochondria isolated from suckling or adult rats [Krauss, S., Lascelles, C.V., Zammit, V.A. & Quant, P.A. (1996) Biochem. J. 319, 427-433]} using a different approach of control analysis, although it is not strictly valid to compare flux control coefficients from different systems. Our overall conclusion is that flux control coefficients for CPT I over oxidative fluxes from palmitate (or palmitoyl-CoA) differ markedly according to (a) the metabolic state, (b) the stage of development, (c) the specific pathway studied and (d) the model system.

Hepatic mitochondrial 3-hydroxy-3-methylglutaryl-coenzyme a synthase deficiency

There are at least two isoenzymes of 3-hydroxy-3-methylglutaryl (HMG)-CoA synthase (EC 4.1.3.5) located in the mitochondrial matrix and the cytoplasm of hepatocytes, respectively. The mitochondrial enzyme is necessary for the synthesis of ketone bodies, which are important fuels during fasting. We report a child with a deficiency of this isoenzyme. He presented at 16 mo with hypoglycemia. There was no rise in ketone bodies during fasting or after a long chain fat load but there was a small rise after a leucine load. Measurement of beta-oxidation flux in fibroblasts was normal. Using antibodies specific for mitochondrial HMG-CoA synthase, no immunoreactive material could be detected on Western blotting. Total HMG-CoA synthase activity in liver homogenate was only slightly lower than in control samples. Presumably, as there was no mitochondrial HMG-CoA synthase enzyme protein, this activity arose from the cytoplasmic or other (e.g. peroxisomal) isoenzymes. With avoidance of fasting, our patient has had no problems since presentation and is developing normally at 4 y of age.

A simple mechanism decreasing free metabolite pool size in static spatial channelling

We propose a simple mechanism which enables decrease of the free pool of channelled metabolite in static spatial channelling, when the concentration of the enzyme consuming the channelled metabolite is greater than the concentration of the enzyme producing this metabolite. Spatial channelling occurs between two enzymes when the common metabolite is released to a small space between these enzymes and does not from a ternary covalent complex with them, as is the case in covalent (dynamic or static) channelling. The mechanism proposed is qualitatively independent of rate constants, metabolite concentrations as well as other kinetic properties and is quantitatively significant for all physiologically relevant conditions. Calculations show that the free metabolite pool must decrease, when the concentration of the enzyme consuming the channelled metabolite is greater than the enzyme producing it. This mechanism is much more effective than increase in the concentration (or rate constant) of the enzyme consuming the metabolite in the absence of spatial channelling.

Flux control exerted by overt carnitine palmitoyltransferase over palmitoyl-CoA oxidation and ketogenesis is lower in suckling than in adult rats

We examined the potential of overt carnitine palmitoyltransferase (CPT I) to control the hepatic catabolism of palmitoyl-CoA in suckling and adult rats, using a conceptually simplified model of fatty acid oxidation and ketogenesis. By applying top-down control analysis, we quantified the control exerted by CPT I over total carbon flux from palmitoyl-CoA to ketone bodies and carbon dioxide. Our results show that in both suckling and adult rat, CPT I exerts very significant control over the pathways under investigation. However, under the sets of conditions we studied, less control is exerted by CPT I over total carbon flux in mitochondria isolated from suckling rats than in those isolated from adult rats. Furthermore the flux control coefficient of CPT I changes with malonyl-CoA concentration and ATP turnover rate.

Regulation and control in complex, dynamic metabolic systems: experimental application of the top-down approaches of metabolic control analysis to fatty acid oxidation and ketogenesis

Metabolic control analysis has provided experimental tools and a precise language to understand and to describe regulation and control quantitatively in complex, dynamic metabolic systems. The top-down approaches of control analysis reduce and simplify the experiments required to analyse: (1) the potential (group control coefficients) of system parameters (blocks of reactions) to control system variables (fluxes, metabolite concentrations); (2) the sensitivity (group elasticity coefficients) of those blocks of reactions to the action of direct and/or indirect effectors and (3) the regulation of system variables, fluxes through blocks of reactions (partial response coefficients) or through the entire pathway (group response coefficients), by internal or external effectors. Although metabolic control analysis expresses regulation and control in terms of response or control coefficients, which are mathematically derived and have numerical values, the two terms are mathematically indistinguishable. In this paper, we define regulation and control as a consequence of hierarchy and not of mathematical function and analyse a conceptually simplified experimental system to demonstrate our distinction.

The role of changes in the sensitivity of hepatic mitochondrial overt carnitine palmitoyltransferase in determining the onset of the ketosis of starvation in the rat

The relationships between the increase in blood ketone-body concentrations and several parameters that can potentially influence the rate of hepatic fatty acid oxidation were studied during progressive starvation (up to 24 h) in the rat in order to discover whether the sensitivity of mitochondrial overt carnitine palmitoyltransferase (CPT I) to malonyl-CoA plays an important part in determining the intrahepatic potential for fatty acid oxidation during the onset of ketogenic conditions. A rapid increase in blood ketone-body concentration occurred between 12 and 16 h of starvation, several hours after the marked fall in hepatic malonyl-CoA and in serum insulin concentrations and doubling of plasma non-esterfied fatty acid (NEFA) concentration. Consequently, both the changes in hepatic malonyl-CoA and serum NEFA preceded the increase in blood ketone-body concentration by several hours. The maximal activity of CPT I increased gradually throughout the 24 h period of starvation, but the increases did not become significant before 18 h of starvation. By contrast, the sensitivity of CPT I to malonyl-CoA and the increase in blood ketone-body concentration followed an identical time course, demonstrating the central importance of this parameter in determining the ketogenic response of the liver to the onset of the starved state.

Flux control exerted by mitochondrial outer membrane carnitine palmitoyltransferase over beta-oxidation, ketogenesis and tricarboxylic acid cycle activity in hepatocytes isolated from rats in different metabolic states

The Flux Control Coefficients of mitochondrial outer membrane carnitine palmitoyltransferase (CPT I) with respect to the overall rates of beta-oxidation, ketogenesis and tricarboxylic acid cycle activity were measured in hepatocytes isolated from rats in different metabolic states (fed, 24 h-starved, starved-refed and starved/insulin-treated). These conditions were chosen because there is controversy as to whether, when significant control ceases to be exerted by CPT I over the rate of fatty oxidation [Moir and Zammit (1994) Trends Biochem. Sci. 19, 313-317], this is transferred to one or more steps proximal to acylcarnitine synthesis (e.g. decreased delivery of fatty acids to the liver) or to the reaction catalysed by mitochondrial 3-hydroxy-3-methyl-glutaryl-CoA synthase [Hegardt (1995) Biochem. Soc. Trans. 23, 486-490]. Therefore isolated hepatocytes were used in the present study to exclude the involvement of changes in the rate of delivery of non-esterified fatty acids (NEFA) to the liver, such as occur in vivo, and to ascertain whether, under conditions of constant supply of NEFA, CPT I retains control over the relevant fluxes of fatty acid oxidation to ketones and carbon dioxide, or whether control is transferred to another (intrahepatocytic) site. The results clearly show that the Flux Control Coefficients of CPT I with respect to overall beta-oxidation and ketogenesis are very high under all conditions investigated, indicating that control is not lost to another intrahepatic site during the metabolic transitions studied. The control of CPT I over tricarboxylic acid cycle activity was always very low. The significance of these findings for the integration of fatty acid and carbohydrate metabolism in the liver is discussed.

Hepatic ketogenesis in newborn pigs is limited by low mitochondrial 3-hydroxy-3-methylglutaryl-CoA synthase activity

In newborn-pig hepatocytes, the rate of oleate oxidation is extremely low, despite a very low malonyl-CoA concentration. By contrast, the sensitivity of carnitine palmitoyltransferase (CPT) I to malonyl-CoA inhibition is high, as suggested by the very low concentration of malonyl-CoA required for 50% inhibition of CPT I (IC50). The rates of oleate oxidation and ketogenesis are respectively 70 and 80% lower in mitochondria isolated from newborn-pig liver than from starved-adult-rat liver mitochondria. Using polarographic measurements, we showed that the oxidation of oleoyl-CoA and palmitoyl-L-carnitine is very low when the acetyl-CoA produced is channelled into the hydroxymethylglutaryl-CoA (HMG-CoA) pathway by addition of malonate. In contrast, the oxidation of the same substrates is high when the acetyl-CoA produced is directed towards the citric acid cycle by addition of malate. We demonstrate that the limitation of ketogenesis in newborn-pig liver is due to a very low amount and activity of mitochondrial HMG-CoA synthase as compared with rat liver mitochondria, and suggest that this could promote the accumulation of acetyl-CoA and/or beta-oxidation products that in turn would decrease the overall rate of fatty acid oxidation in newborn- and adult-pig livers.

A top-down control analysis in isolated rat liver mitochondria: can the 3-hydroxy-3-methylglutaryl-CoA pathway be rate-controlling for ketogenesis?

We incubated isolated liver mitochondria with palmitoyl-CoA, 2,4-dinitrophenol and malonate. Under these conditions all the flux of carbon from palmitoyl-CoA was directed towards acetoacetate synthesis. We measured the rate of acetyl-CoA formation from palmitoyl-CoA (by measuring the rate of oxygen consumption) and the rate of acetoacetate production from acetyl-CoA at three different acetyl-CoA/CoA ratios. Using the top-down approach of metabolic control analysis we calculated the control over ketogenesis exerted by (a) the conversion of extramitochondrial palmitoyl-CoA to intramitochondrial acetyl-CoA and by (b) the conversion of acetyl-CoA to acetoacetate (the 'HMG-CoA pathway'). The overall flux control coefficients of the groups of enzymes involved in (a) and (b) over ketogenesis were 0.28 and 0.72, respectively. Our results show that it is possible for significant control to be exerted over ketogenesis by the enzymes of the HMG-CoA pathway.

Experimental application of top-down control analysis to metabolic systems

Metabolic control analysis (MCA) has provided the language and framework for quantitative study of control over flux, or over metabolites, by individual enzymes of a pathway. By contrast, top-down control analysis (TDCA) yields an immediate overview of the control structure of the whole system of interest, giving information about the control exercised by large sections of complex pathways. Unlike MCA, TDCA does not rely on the use of specific inhibitors or genetic manipulation to determine control coefficients. The method and an application of TDCA to ketogenesis are described.

Control of hepatic mitochondrial 3-hydroxy-3-methylglutaryl-CoA synthase during the foetal/neonatal transition, suckling and weaning in the rat

(1) We assayed active and total (i.e. active plus succinylated) 3-hydroxy-3-methylglutaryl-CoA (HMG-CoA) synthase in mitochondria isolated from foetal, neonatal, suckling or weaned rats. (2) HMG-CoA synthase was substantially succinylated and inactivated in mitochondria isolated from term-foetal, (1-h-old, 6-h-old, 1-day-old) neonatal, suckling and high carbohydrate/low-fat (hc)-weaned rats. Succinylation of HMG-CoA synthase was very low in mitochondria isolated from the livers of foetal, 30-min-old neonatal and high-fat/carbohydrate-free (hf)-weaned rats. (3) There was a negative correlation between active HMG-CoA synthase and succinyl-CoA content in mitochondria isolated from term-foetal, suckling and hc-weaned rats. (4) Differences in active enzyme could not be entirely accounted for by differences in succinylation and inactivation of the synthase. Immunoassay confirmed that the absolute amounts of mitochondrial HMG-CoA synthase increased during the foetal/neonatal transition and decreased with hc weaning. The levels remained elevated with hf weaning. (5) From these data we propose that mitochondrial HMG-CoA synthase is controlled by two different mechanisms in young rats. Regulation by succinylation provides a mechanism for rapid modification of existing enzyme in response to changing metabolic states. Changes in the absolute amounts of HMG-CoA synthase provide a more long-term control in response to nutritional changes.

Expression of hepatic mitochondrial carbonic anhydrase V

We have raised specific (rabbit anti-rat) polyclonal antibodies to hepatic mitochondrial carbonic anhydrase V (CA V) and used them to assay the amounts of protein expressed in liver mitochondria isolated from term-foetal, control or diabetic adult rats and in perivenous and periportal rat hepatocytes. The levels of CA V expressed in mitochondria isolated from the livers of adult male and female rats are similar and increase (about 2-fold) in mitochondria from adult diabetic rats when compared to those isolated from the livers of control rats. The level of enzyme in adult liver was higher than in the livers of term-foetal rats. CA V is expressed in both perivenous and periportal hepatocytes, but the level of expression is greater (approx. 40%) in perivenous cells. The implications and significance of these findings are discussed with reference to the roles and properties of the other carbonic anhydrase isoenzymes and the metabolic function of the mitochondrial isoenzyme.

Glucagon activates mitochondrial 3-hydroxy-3-methylglutaryl-CoA synthase in vivo by decreasing the extent of succinylation of the enzyme

1. 3-Hydroxy-3-methylglutaryl-CoA (HMG-CoA) synthase (EC 4.1.3.5) in extracts of rat liver mitochondria can be inactivated by succinyl-CoA and activated by incubation in a medium designed to cause desuccinylation ('desuccinylation medium'). 2. The enzyme is less active in extracts of whole liver from control rats than from rats treated with glucagon or mannoheptulose. Incubation in desuccinylation medium raises the activity in extracts from control rats to the same value as treated rats, suggesting that the extent of succinylation in vivo is greater in controls than in hormone-treated animals. 3. This result is also obtained in liver homogenates and in isolated mitochondria. 4. Increasing the succinyl-CoA content of mitochondria to the same high level lowers the enzyme activity to the same value in mitochondria isolated from control or treated rats. In each case subsequent incubation of the lysates in desuccinylation medium raises the enzyme activity by the same extent. 5. Measurement of the incorporation of radiolabel from 2-oxo[5-14C]glutarate into protein is consistent with the proposal that all these changes in activity in isolated mitochondria may be explained by changes in the extent of succinylation of the enzyme. 6. From these data and our earlier work we conclude that, in vivo, mitochondrial HMG-CoA synthase in fed rats is normally substantially succinylated (about 40%) and inactivated, and that glucagon increases the activity of HMG-CoA synthase by lowering the concentration of succinyl-CoA and thus decreasing the extent of succinylation of the enzyme (to less than 10%). This may be an important control mechanism in ketogenesis.

Treatment of rats with glucagon or mannoheptulose increases mitochondrial 3-hydroxy-3-methylglutaryl-CoA synthase activity and decreases succinyl-CoA content in liver

1. The activity of 3-hydroxy-3-methylglutaryl-CoA (HMG-CoA) synthase (EC 4.1.3.5) in extracts of rapidly frozen rat livers was doubled in animals treated in various ways to increase ketogenic flux. 2. Some 90% of the activity measured was mitochondrial, and changes in mitochondrial activity dominated changes in total enzyme activity. 3. The elevated HMG-CoA synthase activities persisted throughout the isolation of liver mitochondria. 4. Intramitochondrial succinyl-CoA content was lower in whole liver homogenates and in mitochondria isolated from animals treated with glucagon or mannoheptulose. 5. HMG-CoA synthase activity in mitochondria from both ox and rat liver was negatively correlated with intramitochondrial succinyl-CoA levels when these were manipulated artificially. Under these conditions, the differences between mitochondria from control and hormone-treated rats were abolished. 6. These findings show that glucagon can decrease intramitochondrial succinyl-CoA concentration, and that this in turn can regulate mitochondrial HMG-CoA synthase. They support the hypothesis that the formation of ketone bodies from acetyl-CoA may be regulated by the extent of succinylation of mitochondrial HMG-CoA synthase.