Changes in hepatic fatty acid degradation and blood lipid and ketone body content during development of the rat

The following have been measured during the development of the laboratory rat: the rate of oxidation of palmitoylcarnitine, decanoylcarnitine, and 14C-palmitate by liver mitochondria; the concentrations of ketone bodies in the blood; the plasma concentrations of non-esterified fatty acids, glycerol, and triglyceride. In each case, a rise after birth and a fall at weaning were observed. These changes can be correlated with the dietary changes which occur at these times. However, during the suckling period, when a constant high fat content diet is consumed, further marked changes in the parameters measured were observed which cannot be related to nutritional factors.     

The effect of dexamethasone treatment on the expression of the regulatory genes of ketogenesis in intestine and liver of suckling rats

The influence of the injection of dexamethasone on ketogenesis in 12 day old suckling rats was studied in intestine and liver by determining mRNA levels and enzyme activity of the two genes responsible for regulation of ketogenesis: carnitine palmitoyl transferase I (CPT I) and mitochondrial HMG-CoA synthase. Dexamethasone produced a 2 fold increase in mRNA and activity of CPT I in intestine, but led to a decrease in mit. HMG-CoA synthase. In liver the mRNA levels and activity of both CPT I and mit. HMG-CoA synthase decreased. Comparison of these values with the ketogenic rate of both tissues following dexamethasone treatment suggests that mit. HMG-CoA synthase could be the main gene responsible for the regulation of ketogenesis in suckling rats. The changes produced in serum ketone bodies by dexamethasone, with a profile that is more similar to the ketogenic rate in the liver than that in the intestine, indicate that liver contributes more to ketone body synthesis in suckling rats. Two day treatment with dexamethasone produced no change in mRNA or activity levels for CPT I in liver or intestine. While mRNA levels for mit. HMG-CoA synthase changed little, the enzyme activity is decreased in both tissues.     

Modulation of fatty acid metabolism by glucagon in man. IV. Effects of a physiologic hormone infusion in normal man

This study was done to explore the role of physiologic elevations of glucagon concentration in plasma ketone body concentration in normal man. During the period of hormone elevation, plasma free fatty acids were pharmacologically elevated to ensure adequate free fatty acid substrate delivery to the liver to support hepatic ketogenesis. Eighty-minute infusions of glucagon resulted in a plasma hormone concentration of approximately 300 pg./ml. During the infusion, ketone bodies declined from their basal concentration and remained below basal for the duration of the infusion. An acute heparin-induced pharmacologic elevation of plasma free fatty acid concentration resulted in a transient rise in plasma ketone body concentration, but at no time did it attain the concentration observed during the control saline infusion. Plasma glucose concentration was not altered by glucagon infusion, but plasma insulin concentration rose by approximately 2.5 muU./ml. These results suggest that glucagon is not ketogenic in normal man as has been previously reported in insulin-deficient diabetics. The glucagon-induced rise in plasma insulin concentration may participate in the observed reduction in plasma ketone body concentration.     

The value of beta-2 transferrin analysis and MR cisternography for the diagnosis of cerebrospinal fluid fistulas

Blood chemical values, including ketone bodies, were measured in 25 cows with abomasal displacement (displacement group), 16 cows with primary ketosis (ketosis group), and nine normal controls to investigate the pathophysiology of abomasal displacement. Increases in aspartate aminotransferase, gamma-glutamyl transpeptidase, non-esterified fatty acid (NEFA), and ketone bodies (3-hydroxybutyric acid and acetoacetic acid) were observed in the displacement and ketosis groups. Total cholesterol increased significantly in the ketosis group but decreased in the displacement group. Glucose was significantly low and reversely correlated to ketone bodies in the ketosis group but was not low and was not correlated with ketone bodies in the displacement group. While NEFA was correlated to ketone bodies in the ketosis group, it was not in the displacement group. A correlation between the values of acetoacetic acid and 3-hydroxybutyric acid was seen in both the ketosis and displacement groups. The fact that blood chemical values in ketosis cows were clearly different from those in displacement cows suggest that the biochemical mechanism of ketogenesis is different between these two groups.     

Oxygen consumption and biosynthetic function in perfused liver from rats at different stages of development

Changes in mitochondrial function were studied in perfused liver from rats aged 24-365 days. Oxygen consumption together with the rates of gluconeogenesis, urea synthesis and ketogenesis were determined. Basal mitochondrial respiration as well as the ability of the liver to synthesize glucose, urea and ketone bodies declined from 24- to 365-day-old rats. On the other hand, on transition from 24 to 60 days the liver oxidation rate of hexanoate, sorbitol and glycerol is enhanced, but not of ketone bodies or palmitate. Our results show that the transition from weaning to middle age is accompanied by defined changes in hepatic substrate oxidation. From the observed time course of the decrease in basal and substrate-stimulated oxygen consumption, it is concluded that in rat liver cells a decline in respiratory chain function, long-chain fatty acid and ketone body metabolism, gluconeogenesis and ureogenesis occurs at a relatively early life stage.     

Plasma lipids, ketone bodies, and glucose concentrations in calves fed high- and low-fat milk replacers

Twenty-four 5-day-old male calves were fed twice daily milk replacers containing either 5% (low-fat) or 25% (high-fat) lard. Plasma lipids, blood glucose, and ketone bodies were determined in jugular blood before feeding and every hour during 8 h after feeding. The high-fat diet caused in the 1st h after feeding a sharp increase of triglycerides and phospholipids followed by a sharp decrease; these two increased slowly during the following 5 h. Within the first 2 h after feeding, there was an increase of cholesterol esters, free cholesterol, and nonesterified fatty acids. With the low-fat diet, triglycerides and cholesterol esters showed a small increase during the 4 h following meal whereas phospholipids, free cholesterol, and nonesterified fatty acids were not affected significantly. With both diets, blood glucose reached a maximum of 110 mg/100ml 1 h after feeding; ketone bodies were not altered significantly. With the high-fat diet, lipid digestion would occur in two phases; firstly, part of the fat would be lipolyzed quickly by pregastric esterase before clot formation in the abomasum; secondly, the rest of the lipids, slowly released by progressive lysis of the coagulum would be digested under the action of gastric and pancreatic lipases. The first phase did not occur with the low-fat diet.     

Lipolytic enzyme and phospholipid level changes in intraoperative salvaged blood

Autotransfusion is becoming increasingly popular, mainly because it eliminates the risk of disease transmission. One of the techniques available is intra-operative blood salvage and retransfusion with or without washing of the collected blood. The blood collected during this process is subjected to a variety of chemical and physical insults which can alter the normal composition of the plasma by activating plasma and cellular homeostatic mechanisms. In this study, we measured the plasma levels of total phospholipids, lysolecithin and non-esterified fatty acids, and the lipolytic enzymes phospholipase A2 (PLA2) and lipase in the salvaged blood before and after washing. In the unwashed salvaged blood the mean levels of PLA2, non-esterified fatty acids and lysophospholipids increased by 144, 96 and 149%, respectively, while those of total phospholipids and lipase did not change to any extent. All these substances were reduced to well below the patients circulating plasma levels by washing the collected blood. The changes indicate that the lipid profile of salvaged blood is significantly altered and that potentially dangerous substances such as PLA2 and its metabolites, lysolecithin and non-esterified fatty acids, are present in increased amounts. Washing the blood is recommended prior to reinfusion.     

On the mechanism of regulation of omega oxidation of fatty acids

The stimulatory effect of starvation on omega oxidation of stearate by the 20,000 X g supernatant fluid of rat liver homogenates was studied. The effect was obtained after starvation for 24 hours. Starvation for longer times did not further increase omega oxidation. The stimulatory effect of starvation on omega oxidation of stearic acid was accompanied by a reduced incorporation of stearic acid into phosphatidic acid, diglycerides, and triglycerides. Substitution of the 100,000 X g supernatant fluid from liver homogenate of starved rats with 100,000 X g supernatant fluid from liver homogenates of control rats reduced the microsomal omega oxidation of stearic acid with a simultaneous increase in incorporation of stearic acid into the different glycerides. Under the latter conditions almost no free stearic acid could be isolated from the incubation mixture after the incubation. Of three different soluble factors necessary for glyceride formation, ATP appeared to be the most important from a regulatory point of view. Thus the soluble fraction of liver homogenate from a starved rat was shown to contain suboptimal concentrations of ATP. Addition of physiological amounts of ATP to the 20,000 X g supernatant fluid of homogenate of liver of starved rats had the same effect as addition of 100, 000 X g supernatant fluid from liver homogenate of control rats, i.e. decrease in omega oxidation and increase in formation of glycerides. Addition of sn-glycerol 3-phosphate and CoA-SH in amounts optimal for glyceride formation to the 20,000 X g supernatant fluid of liver homogenate of starved rats had only small effects on omega oxidation and glyceride formation. The results are consistent with a competition for free fatty acids between the acyl-CoA synthetases involved in biosynthesis of glycerides and the microsomal hydroxylase(s) involved in omega oxidation of fatty acids. The concentration of ATP in the soluble fraction is of importance in this competition. The possibility is discussed that this competition is of importance also under in vivo conditions and that a decreased rate of esterification in the starved state is responsible for the higher excretion of omega-oxidized fatty acids in urine in the ketotic state.     

Thermoregulation in newborn lambs: influence of feeding and ambient temperature on brown adipose tissue

We have previously shown that feeding 50 ml of colostrum can increase the thermogenic activity of brown adipose tissue (BAT) in newborn lambs maintained at a warm (30 degrees C) ambient temperature. This study further examines the effect of ambient temperature on BAT and thermoregulation by investigating the response to feeding 50 ml of water. Immediately after vaginal birth, lambs were placed in either a warm (30 degrees C) or cool (15 degrees C) environment a ambient temperature and measurements of colonic temperature and heat production were recorded for 6 h. Lambs were fed 50 ml of water when 5 h old. The level of guanosine 5'-diphosphate (GDP) binding was higher, but adrenaline content lower in BAT sampled from lambs maintained at 15 degrees C compared with those at 30 degrees C. Feeding was associated with an increase in colonic temperature and plasma concentrations of glucose and non-esterified fatty acids in lambs maintained at 15 degrees C only. In this group plasma concentrations of adrenaline and dopamine declined after feeding, but noradrenaline concentrations were not influenced by feeding in either group of lambs. O2 consumption and CO2 production were higher in lambs maintained at 15 degrees C but were not influenced by ambient temperature or feeding. It is concluded that feeding a small volume of water can influence thermoregulation by a mechanism that is dependent on the ambient temperature at which the lamb is maintained.     

Clinical, haematological, metabolic and endocrine traits during the first three months of life of suckling simmentaler calves held in a cow-calf operation

Newborn suckling Simmentaler calves (10 males and 9 females) in a cow-calf operation were examined from birth up to the age of 3 months. The average daily gain from 47 to 120 kg was 0.86 kg. Except for higher average daily weight gains and insulin-like growth factor-I concentrations and lower thyroid hormone levels in male than female calves, there were no significant sex differences. Plasma glucose, total protein and immunoglobulin G concentrations increased on day 1 of life, thrombocyte number and plasma triglyceride concentrations rose during the first 7 days, whereas lymphocyte and monocyte percentage and plasma inorganic phosphorus, phospholipid, cholesterol and albumin concentrations increased during the first 14 or 21 days and then remained elevated. Eosinophil percentage increased after 3 weeks and insulin-like growth factor-I concentrations increased over the whole growth period. There were transient elevations of plasma glucagon concentrations up to day 14, of the activity of alkaline phosphatase transiently up to day 7 and of gamma-glutamyltransferase, aspartate aminotransferase and lactate dehydrogenase activities on day 1 of life. Plasma iron concentration transiently decreased up to day 28 and creatine kinase activity up to day 7. Total white blood cell number, neutrophil percentage, packed cell volume and concentrations of haemoglobin, calcium, magnesium (after a transient rise on day 1), non-esterified fatty acids, bilirubin, creatinine, triiodothyronine and thyroxine decreased from birth up to days 42, 56, 28, 28, 21, 84, 14, 14, 7, 14 and 7, respectively. Basophil percentage and concentrations of beta-hydroxybutyrate, urea and insulin did not exhibit significant age-dependent changes. The behaviour of most traits in the first weeks was the same in suckling calves under study as in non-suckling pre-ruminant calves. However, packed cell volume, red blood cell number, haemoglobin and plasma iron concentrations were higher, whereas glucose and insulin concentrations were lower than normally found in veal calves. On the other hand, concentrations of glucose, insulin and insulin-like growth factor-I in suckling calves in the third month of age were higher than can normally be measured in breeding calves.     

Weanling and adult rats differ in fatty acid and carnitine metabolism during sepsis

Increased oxidation of fat is an important host response to sepsis, and carnitine is essential for long-chain fatty acid oxidation. Because neonates have low levels of carnitine, their ability to respond to a septic insult may be impaired. The purpose of this study was to compare fatty acid and carnitine metabolism in septic weanling (60 to 85 g) and septic adult (285 to 310 g) rats. Sepsis was induced in weanling and adult male Sprague-Dawley rats by cecal ligation and puncture (CLP). The rats were killed 16 hours after CLP or sham operation, and serum glucose, lactate, beta-hydroxybutyrate, fatty acid, carnitine, liver fatty acid, and tissue carnitine levels were measured. The data suggest that during sepsis weanling rats may be more dependent on fatty acid oxidation than adult rats are, as evidenced by their elevated serum fatty acid and acylcarnitine levels, and relative hypoglycemia and hyperketonemia. In addition, although total serum carnitine levels were increased in both adult and weanling septic rats, tissue carnitine levels of weanling rats became significantly depleted during sepsis, unlike in adult rats. This study supports further investigation regarding the role of exogenous carnitine in newborn sepsis.     

Early modulation of genes encoding peroxisomal and mitochondrial beta-oxidation enzymes by 3-thia fatty acids

The aim of the present study was to elucidate the effects of a single dose of 3-thia fatty acids (tetradecylthioacetic acid and 3-thiadicarboxylic acid) over a 24-hr study period on the expression of genes related to peroxisomal and mitochondrial beta-oxidation in liver of rats. The plasma triglyceride level decreased at 2-4 hr, 4-8 hr, and 8-24 hr, respectively, after a single dose of 150, 300, or 500 mg of 3-thia fatty acids/kg body weight. Four to eight hours after administration of 3-thia fatty acids, a several-fold-induced gene expression of peroxisomal multifunctional protein, fatty acyl-CoA oxidase (EC 1.3.3.6), fatty acid binding protein, and 2,4-dienoyl-CoA reductase (EC 1.3.1.43) resulted, concomitant with increased activity of 2,4-dienoyl-CoA reductase and fatty acyl-CoA oxidase. The expression of carnitine palmitoyltransferase-I and carnitine palmitoyltransferase-II increased at 2 and 4 hr, respectively, although at a smaller scale. In cultured hepatocytes, 3-thia fatty acids stimulated fatty acid oxidation after 4 hr, and this was both L-carnitine- and L-aminocarnitine-sensitive. The hepatic content of eicosapentaenoic acid and docosahexaenoic acid decreased throughout the study period. In contrast, the hepatic content of oleic acid tended to increase after 24 hr and was significantly increased after repeated administration of 3-thia fatty acids. Similarly, the expression of delta9-desaturase was unchanged during the 24-hr study, but increased after feeding for 5 days. To conclude, carnitine palmitoyltransferase-I expression seemed to be induced earlier than 2,4-dienoyl-CoA reductase and fatty acid binding protein, and not later than the peroxisomal fatty acyl-CoA oxidase. The expression of delta9-desaturase showed a more delayed response.     

Effects of exogenous hormones and glucose on plasma levels and hepatic metabolism of amino acids in the fetus and in the newborn rat

The present study examines the role of insulin, glucagon and cortisol in the regulation of gluconeogenesis from lactate and amino acids in fetal and newborn rats. Injection of glucagon in the full-term fetal rat caused a rise in glucose (and insulin) and a fall in blood levels of most individual amino acids, stimulated hepatic accumulation of 14C-amino isobutyric acid and 14C-cycloleucine and increased the conversion of 14C lactate, alanine and serine to glucose in vivo and in vitro (liver slices). Such changes were equivalent to the changes seen in 4 h old newborn rats. When glucagon was administered at birth, little difference was observed between control and treated animals in plasma amino acids and a smaller increment in conversion of 14C substrate to glucose occurred. By contrast, insulin injection at birth caused hypoglycemia, suppression of levels of certain amino acids and inhibition of conversion of 14C substrates into glucose. Glucose injection at birth caused elevated glycemia and plasma insulin and suppression of most amino acid levels and of conversion of 14C substrate into glucose. Cortisol injection at birth caused a marked, generalized by hyperaminoacidemia, a stimulation of glucagon secretion and of conversion of 14C substrates into glucose. These observations support the thesis that glucagon plays a major role in the induction of hepatic gluconeogenesis and that insulin acts as an antagonist hormone.     

Liver and heart mitochondrial succinate dehydrogenase activity of newborn rats in anoxic hypoxia and starvation

Succinate dehydrogenase activity was determined in the liver and heart of newborn rats after 3 and 48 hours' exposure to anoxic hypoxia (10% O2) and after 48 hours' starvation. Control determinations were made on newborn animals of corresponding ages, full term foetuses (21 days), infantile (1 and 2 weeks) and full grown animals. Hypoxia for 3 h had no influence on succinate dehydrogenase activity at all in either the heart or liver mitochondria of the newborn animals. After 48 h no difference was observed in the liver between the hypoxic animals and the starved controls of the same age, though starvation itself had resulted in a significant increase in activity, as much as 42%. When liver mitochondrial succinate dehydrogenase in normal mitochondria was activated by preincubation mitochondria with the substrate, the activity increase obtained was greater than that resulting from starvation. The increase in activity in the heart of the hypoxic or starved animals was not significant (less than 10%).     

In vitro inhibition of carnitine acyltransferase activity in mitochondria from rat and mouse liver by a diethylhexylphthalate metabolite

The effects of mono(2-ethyl-5-oxohexyl)phthalate [ME(O)HP], a di(2-ethylhexyl)phthalate (DEHP) metabolite and a potent peroxisomal inducer, on the mitochondrial beta-oxidation were investigated. In isolated rat hepatocytes, ME(O)HP inhibited long chain fatty acid oxidation and had no effect on the ketogenesis of short chain fatty acids, suggesting that the inhibition occurred at the site of carnitine-dependent transport across the mitochondrial inner membrane. In rat liver mitochondria, ME(O)HP inhibited carnitine acyltransferase I (CAT I; EC 2.3.1.21) competitively with the substrates palmitoyl-CoA and octanoyl-CoA. An analogous treatment of mouse mitochondria produced a similar competitive inhibition of palmitoyl-CoA transport whereas ME(O)HP exposure with guinea pig and human liver mitochondria revealed little or no effect. The addition of clofibric acid, nafenopin or methylclofenopate revealed no direct effects upon CAT I activity. Inhibition of transferase activity by ME(O)HP was reversed in mitochondria which had been solubilized with octyl glucoside to expose the latent form of carnitine acyltransferase (CAT II), suggesting that the inhibition was specific for CAT I. Our results demonstrate that in vitro ME(O)HP inhibits fatty acid oxidation in rat liver at the site of transport across the mitochondrial inner membrane with a marked species difference and support the idea that induction of peroxisome proliferation could be due to an initial biochemical lesion of the fatty acid metabolism.     

Effect of fluoxetine on the plasma concentrations of clozapine and its major metabolites in patients with schizophrenia

Triglyceride levels and free fatty acid metabolism are influenced by body fat distribution. To test whether the pattern of fat distribution in obese subjects results in distinct insulin mediated suppression of non-esterified fatty acids which could account for differences in plasma triglycerides, we studied 59 obese subjects who were classified according to waist-to-hip ratio. Non-esterified fatty acids and insulin response to a 75 g oral glucose tolerance test were higher in abdominal obesity. Total non-esterified fatty acids response, after adjustment for sex, showed a positive association with waist-to hip ratio (r = 0.292; p < 0.05). The abdominal obese group had higher fasting triglycerides (1.74+/-0.83 versus 1.11+/-0.71 mmol/L; p = 0.003) and lower glucose/insulin ratio (5.2+/-2.3 versus 7.1+/-2.4; p = 0.003). Stepwise multiple regression analysis showed that triglyceride levels are explained by fasting and 120 min non-esterified fatty acids and by glucose/insulin ratio. We conclude that abdominal obesity is associated with a higher resistance to insulin mediated suppression of non-esterified fatty acids in obese subjects. Variation of triglyceride concentrations in obesity is dependent on both fasting and 120 min non-esterified fatty acids as well as on insulin sensitivity to glucose utilization.     

Active transport of butyrobetaine and carnitine into isolated liver cells

1. The liver cells lose the major part of their carnitine during the commonly used isolation procedure by the collagenase-perfusion method. 2. The cells take up carnitine and the carnitine precursor butyrobetaine when these substances are added to the medium. The carnitine content of isolated liver cells can increase to about 15 mM with no apparent harm to the cells. 3. The data indicate the existence of a common carrier in the plasma membrane which mediates the uphill transport of both carnitine and butyrobetaine. The carrier has a high affinity for butyrobetaine (Km=0.5 mM) and a lower one for carnitine (Km=5.6 mM). 4. The intracellular butyrobetaine is hydroxylated to carnitine with a rate of approximately 0.33 mumol-g wet weight-1-h-1 which is sufficient to cover the turn over of carnitine in the whole rat. Carnitine is effectively esterified in the liver cells to acetylcarnitine and long-chain acylcarnitines. 5. Both carnitine and acetylcarnitine are released from the cells. The release of both compounds is probably physiological since it was found that acetylcarnitine constitutes a similar fraction of the total acid soluble carnitine both in the blood and liver of the intact rat.     

Modifying the n-3 fatty acid content of the maternal diet to determine the requirements of the fetal and suckling rat

During perinatal development, docosahexaenoic acid (22:6n-3) accumulates extensively in membrane phospholipids of the nervous system. To evaluate the n-3 fatty acid requirements of fetal and suckling rats, we investigated the accumulation of 22:6n-3 in the brain and liver of pup rats from birth to day 14 postpartum when their dams received increasing amounts of dietary 18:3n-3 (from 5 to 800 mg/100 g diet) during the pregnancy-lactation period. The fatty acid composition of brain and liver phospholipids of pups, as well as that of dam's milk, was determined. At birth, 22:6n-3 increased regularly to reach the highest level when the maternal diet contained 800 mg 18:3n-3/100 g. On days 7 and 14 postpartum, brain 22:6n-3 plateaued at a maternal dietary supply of 200 mg/100 g. Docosapentaenoic acid (22:5n-6) had the opposite temporal pattern. The unusually high concentration of eicosapentaenoic acid (20:5n-3) in liver and dam's milk observed at the highest 18:3n-3 intake suggests an excessive dietary supply of this fatty acid. All these data suggest that the n-3 fatty acid requirements of the pregnant rat are around 400 mg 18:3n-3 and those of the lactating rat at 200 mg (i.e., 0.9 and 0.45% of dietary energy, respectively). The values of 18:3n-3 and 22:6n-3 milk content which allowed brain 22:6n-3 to reach a plateau value in suckling pups were 1% of total fatty acids and 0.9% (colostrum) to 0.2% (mature milk), respectively. These levels are similar to those recommended for infant formulas.     

Cerebral utilization of glucose, ketone bodies and oxygen in starving infant rats and the effect of intrauterine growth retardation

Cerebral arteriovenous differences of acetoacetate, D-beta-hydroxybutyrate, glucose, lactate and oxygen and brain DNA content was measured at 20 days of age in intrauterine growth retarded (IUGR) rats and normal littermates after 48 and 72 h of starvation. Cerebral blood flow (CBF) was measured with labeled microspheres in other comparable groups of IUGR and control rats. CBF was similar in IUGR and normal littermates (0.57+/-0.09 and 0.58+/-0.10 ml/min respectively). After 48 h of starvation, arterial glucose was significantly lower in IUGR than control animals but the arterial concentrations of ketone bodies were similar. After 48 h of starvation, cerebral arteriovenous difference of beta-hydroxybutyrate was significantly higher in control than IUGR rats also when expressed per mg brain DNA as was the fractional uptake of D-beta-hydroxybutyrate. After 72 h of starvation, arterial concentrations of ketone bodies were significantly lower in IUGR rats than controls but the fractional uptake of D-beta-hydroxybutyrate was increased compared to IUGR rats starved for 48 h. The average percentage of calculated total substrate uptake (mumol/min) accounted for by ketone bodies increased in control animals from 31.1% after 48 h of starvation to 41.0% after 72 h of starvation. In IUGR rats these percentage values were 26.5 and 25.7 respectively. After 72 h of starvation the fraction of total cerebral uptake of substrates accounted for by ketone bodies was significantly higher in control that IUGR rats. As total cerebral uptake of substrates was similar between IUGR and control animals it is concluded that IUGR rats are more dependent on glucose as a substrate for the brain during starvation.