Impaired ketone body metabolism in the selenium deficient rat. Possible implications

Impaired Ketone Body Metabolism in the Selenium Possible Implications.

Deficient

Rat.

Ulf Olsson Male rats were fed a selenium-deficient Torula yeast diet with or without 0.2 ppm selenium (as sodium selenite) in the drinking water. Selenium deficiency caused a significant increase of urinary acetoacetate excretion in fed rats, and 24 or 45 hours of starvation enhanced this effect. Two days of selenium supplementation decreased the amount of urinary acetoacetate and 3-hydroxybutyrate to 50% of the deficiency value, indicating an enzymatic impairment in the selenium-deficient rat. No selenium-dependent effect was found for the following: (1) urinary pH, amount of nitrite, glucose (2) blood content of glucose, (negative), hemoglobin or protein, and the urine was negative for phenylketones; acetoacetate, or 3-hydroxybutyrate; or (3) liver content of glycogen, glucose, acetoacetate. or 3-hydroxybutyrate. On the other hand, the liver content of triglycerides was significantly lower in selenium deficiency. Indications for a higher content of ketone bodies (acetoacetate plus 3-hydroxybutyrate) in the kidneys from selenium-deficient rats were found. The increased urinary excretion of ketone bodies on selenium deficiency may indicate an impairment of lipid and ketone body turnover (in the kidney), or a decreased kidney reabsorption rate. Possible implications of these results in connection with protective roles of selenium in atherosclerosis and carcinogenesis are suggested. D 1985 by Grune & Stratton, Inc.

I

(Se) was identified as the essential component in factor 3 found to prevent dietary liver necrosis in vitamin E-deficient rats.’ One important function of selenium was later found to be as a cofactor in the enzyme glutathione peroxidase (GSHPx), for a review see ref 2. Most work on biologic functions of selenium has focused on the protective role of this enzyme in removing (lipid) peroxides.3 A connection between selenium deficiency and a wide range of diseases in animals4a and humans4b has recently been reviewed. Human cancer& and chemically induced tumor frequencies,5 as well as some cardiovascular disease&’ have been reported or suspected to correlate with selenium deficiency. However, the physiologic functions of selenium in these different states of diseases are not fully understood, and loss of GSH-Px activity is not enough to explain the connection between selenium deficiency and these diseases or malfunctions. Some possible functions of selenium, not related to GSH-Px, have been reviewed.’ Such functions may include a protective effect of selenium against diquat toxicity and a role to stabilize heme catabolism.8 In this report, an impaired ketone body metabolism in the Se-deficient rat is reported, an effect mainly manifested as a highly increased urinary excretion of acetoacetate ( AcAc) and 3-hydroxybutyrate (3-OHBA) on starvation. In a computer search of the literature from 1969 to October 1983, no report of a connection between selenium and ketone body metabolism was found, and the mechanism behind the present finding is unknown. However, the possible importance of selenium in ketone body metabolism, and of ketone bodies in explaining some selenium protective effects is briefly discussed with the purpose of stimulating further research on this subject.

N 1957 selenium

Merabolism, Vol 34, No

11 (November),

1985

MATERIALS

AND

METHODS

Animals Male Wistar rats, originally from Banting & Kingman Ltd (Grimstone Hull, England), were used. They were fed a standard diet (Ewos R3) or one of two partly different Se-deficient diets* (Torula I and II). Selenium content (dry basis) of the standard and Torula diet was 0.22 and 0.003 ppm respectively, as determined by neutron activation analysis. This analysis was performed by the Swedish Environmental Research Institute, Institutet for Vattenoch Luftvardstorskning (IVL). Tap water with less than 0.001 ppm of selenium was given to the standard diet groups, while distilled water with SO mg sodium chloride per liter was given to the Torula diet groups. Selenium supplementation was with 0.2 ppm Se (as Na,SeO,) in the drinking water. In some cases the rats were fed diet I with or without selenium supplementation from 4 to 5 weeks of age. Other groups of rats were born from dams fed diet I or II for 5 weeks before partus. This offspring either was or was not supplemented with selenium from 5 weeks of age. The rats were matched, in that each Se-deficient rat

*Composition of Torula diets I and II: Torula yeast 30%, sucrose 43.8%. corn starch 15%,vitamins 1% (Burk RF et al, J Nutr 95:420. 1968, except that 320 IU vitamin D, and 5.0 mg vitamin E acetate was added per 100 g of the final diet) and dl-methionine 0.3%. Lipid and salt mixture differed between diets. Diet I contained tocopherolstripped corn oil 6.4% and William-Briggs modified salt mixture 3.5% (J Nutr 103:536, 1973) while diet II contained tocopherolstripped corn oil 3.2% and lard 3.2%. and Hubbel-MendelWakeman salt mixture 3.5% (J Nutr 15:273, 1937). All ingredients. except sucrose and corn starch, were purchased from the United States Biochemical Corp, Cleveland. From the Division of Cellular Toxicology, Wallenberg Laboratory, University of Stockholm. Supported by the National Swedish Environmental Protection Board and the Swedish Natural Science Research Council. Address reprint requests to Dr Olsson, Division of Cellular Toxicology, Environmental Toxicology Unit, Wallenberg Laboratory, University of Stockholm, S-106 91 Stockholm, Sweden. 0 1985 by Grune & Stratton, Inc. 0026-0495/85/3411-0002%03.00/0

993

994

ULF OLSSON

corresponded to a Se-supplemented sibling. They were 20 to 25 weeks old, and selenium supplementation was for 15 to 20 weeks, except for the repletion study in which supplementation for two, seven, and 12 days was also used. Sample Collection All rats were held in metabolic cages for 4 to 5 hours (fed) or 24, 48, and 72 hours (starved) to collect the urine. The rats were stunned by a blow on the head and decapitated. Blood was collected from the neck to chilled test-tubes and frozen in an ethanol-(carbon dioxideice) mixture. Livers were removed as quickly as possible, put between two sheets of aluminum foil, and frozen between two planed blocks of carbon dioxide-ice. The time between decapitation and start of freezing was 25 to 35 seconds. Livers were stored at -70 “C for not more than one month. Analysis Glutathione peroxidase assay was by a modified coupling method of Paglia and Valentine,9 as described by Gtinzler et al” with 0.25 mmol/L hydrogen peroxide as the substrate, and the 105,000 g supernatant from liver was the enzyme source. Unit activity is expressed as rmol NADPH oxidized/min/g protein. The protein content was determined according to Lowry” using bovine albumin as standard. Liver samples were used to determine triglyceride,” glycogen, and glucose” content. Blood glucose determination was according to Bergmeyer et al.” AcAc15 and 3-0HBA16 were determined in liver, blood, and urine. In initial experiments, the urine was tested for pH, nitrite, glucose,

60

1

hemoglobin, protein, and AcAc using Ames N-labstix 28 15. Subsequently, the urine was only tested for AcAc with ketostix 2880K, and diluted with distilled water before enzymatic analysis was performed on fresh or frozen (-70 “C) samples. Freezing did not affect the value. For the enzymatic analysis of ketone bodies (AcAc and 3-OHBA) in urine, a dilution with distilled water of not less than I:20 was required to obtain a linear correlation between dilution and extinction. Values for urinary AcAc determined with ketostix 2880K and enzymatically’ agreed well up to about 1 pmol/mL. If the urine contained more AcAc, the ketostix underestimated the value. However, only enzymatically determined values for the content of ketone bodies are presented. Data on the urinary content of ketone bodies are presented as amount excreted per unit time. However, the results could have been expressed per milliliter of urine without changing the significant differences between Se-deficient and Se-supplemented rats. The data obtained were tested statistically using Student’s t test. Comments

on Analysis of Ketone Bodies

The values for liver, kidney, and blood ketone bodies are low compared to those reported in the literature.” This might depend on the freezing technique adopted. A freeze clamp technique” using liquid nitrogen and faster removal of organs, is preferable to measure absolute or near in vivo values of some tissue metabolites, such as ketone bodies.19 However, the present and simpler technique of freezing was regarded as adequate to estimate relative ketone body concentrations. Therefore, these values should be taken as relative comparisons between Se-deficient and Se-supplemented groups, rather than as absolute values. In the original experimental series urinary 3-OHBA was not determined. This was performed at a later date, when the frozen (- 70 “C) urine samples were reexamined for both ketones. The loss of AcAc during storage was about 6% (6 k 3%. mean -t SE of 12 samples) and the reexamined values were corrected for this loss (Fig 1 and Table 1). This slow decomposition rate of AcAc was apparently due to the storage condition (- 70 “C). Harano et alz9 reported a 50% decomposition of AcAc after 2 weeks at - 20 “C, while no decrease was noted after 3 weeks at - 70 “C. 3-OHBA was stable on storage at -20 “C or -70 0C2’ Chemicals The highest purity chemicals were purchased from Sigma Co (St Louis). The stix were from Ames Division, Miles Laboratories Limited, Stoke Poges, Slough SL2 4LY, England. Table 1. Effect of Selenium Deficiency on Glutethione Peroxidase (GSH-Px) Activity and Triglyceride Content of Liver and on the Urinary Content of Ketone Bodies in Rats Starved for 48 Hours

Diet Regimen Se-supplemented

I.. 0

Se-deficient 2 Days of selenium

7

12

97

supplementation

Fig 1. Effect of selenium supplementation to deficient rats on the urinary excretion of ketone bodies. Rats were held in metabolic cages for 48 hours and the urine from 24 to 48 hours of starvation was collected and analyzed for 3-hydroxybutyrete (0) and acatoacetate (0). For a full description, see Results. Values are given as means + SE of 4 rats. Significantly different from unsupplemented rats are indicated as lP < 0.05 or l*P c 0.01.

KetoneBcdiest

GSH-Px*

TG (mg TG/g liver)

186 + 10

22 + 2

5*1

6+2

11 * 1

38+2

ACAC

J-CHBA 6*1 41-c8

Rats ware held in metabolic cages and starved for 48 hours. Values era means + SE of 12 (GSH-Pxand TG) and 8 (ketone bodies) rats. For the 4 parameters determined, P -c 0.00

1 for the Se-dependent difference.

lGSH-Px activityis expressed in units, U = pmol NADPH oxidized/ min/g protein of liver 105.000

g supernatant.

tpmol acetoacetate (AcAc) and 3-hydroxybutyrate (3-OHEA) per 24 hours urine. Urine collected during the second 24 hours of a 4%hour period of starvation. The urine volume was 14 + 4 and 9.9 * 2.5 mL for Se-deficient and Se-supplemented rats, respectively.

IMPAIRED

KETONE BODY METABOLISM

RESULTS

Irrespective of diet or breeding regimen (see Materials and Methods) the rats were very similar in the response to selenium and starvation. Therefore, the results are presented only in terms of selenium deficiency or selenium supplementation. Se-supplemented groups, and rats fed the standard diet did not differ from each other in their urinary excretion of AcAc, neither when starved nor during any of the different times of starvation that were tested (not shown). Beside the enzymatic determinations of ketone bodies in urine and tissues from Se-deficient and Sesupplemented rats the urine was also tested with labstix 2815. The tests with labstix showed no Sedependent difference in urinary pH, nitrite, glucose, hemoglobin, or protein excretion. When Se-supplemented rats were placed in metabolic cages with and without selenium in the drinking water, no difference in urinary AcAc content was found between these two groups, thus excluding a false-negative result for the Se-supplemented group.

Effect of Starvation

on Urinary AcAc Content

The AcAc content in urine from Se-deficient and Se-supplemented rats was determined after 24,48, and 72 hours of starvation (Fig 2). Excretion of AcAc increased on starvation and reached a maximum at about 48 hours. The total amount of AcAc excreted per day revealed an extremely high value for one Se-deficient rat. This high value depended on a higher content of AcAc per milliliter of urine, as well as on a higher urine volume than for the other Se-deficient rats. Interestingly, high-excreting rats (from this and other experiments) were from the same litters as Se-supplemented controls who, in turn, showed a lower excretion of AcAc than other litters. This indicates some genetic variation, with some litters being very dependent on selenium in order to maintain a “normal” ketone body metabolism. However, the data on this observation are too sparse to be statistically verified.

Decrease of Urinary Ketone Bodies on Selenium Supplementation

to Deficient

Rats

The urinary excretion of AcAc and 3-OHBA in Se-deficient rats was decreased to about 50% after two days of selenium supplementation (Fig 1). This figure represents an experiment where the same four rats were repeatedly starved for 48 hours and the urine collected during the latter half (24-48 hours) of the starvation period and analyzed for AcAc and 3-OHBA content. The rats were given nonSe-supplemented

Fig 2. Effect of selenium deficiency and of starvation on the urinary excretion of acetoacetate. Se-deficient (0) and Se-supplemented (0) rats were held for four hours funstarved) and up to 72 hours (starved) in metebolic cages. The urinary content of acetoacetate is expressed per 24 hours urine, except for the unstarved values which are for four hours. Values are means + SE of 4 rats, except when one high-excreting animal was excluded. If not indicated, the SE bar is smaller than the symbol. Significant increase on selenium deficiency is indicated as lP < 0.05, l*P < 0.01.0r~**P~0.001.

drinking water when starved and each starvation period was followed by five days refeeding, of which the last two or all five days were with selenium supplementation, thereby giving two, seven, and 12 days of selenium supplementation. The group that had been supplemented for 97 days was from the same litters as the other group, but Se-supplemented from an age of 5 weeks, and starved only once. Both groups were about 20 weeks old. Note that the variance for the AcAc values (Fig 1) decreased with prolonged selenium supplementation, and also that seven and 12 days of supplementation were not significantly different from the group supplemented for 97 days. The value for 12 days selenium supplementation is probably an overestimate as it was noted in other experiments that more than three periods with 48 hours starvation and five days refeeding led to a nonsignificant increase of AcAc excretion (not shown). This may indicate incomplete restoration of the liver glycogen storage after more than three cycles of starvation-refeeding. There was thus a rapid decrease in urinary AcAc and 3-OHBA excretion on selenium supplementation, and the significance was the same if expressed per 24 hours urine (Fig 1) or per milliliter of urine (not shown). Selenium supplementation caused a nonsignif-

ULF OLSSON

icant increase shown).

of the 3-OHBA/AcAc

ratio

(not

Ketone Bodies of Urine, Blood, and Liver, in Relation to the Main Hepatic Energy Source Starvation is followed by depletion of liver glycogen and a decreased supply of glucose to peripheral tissues. This initiates ketone body production in the liver by mobilization of triglycerides and subsequent oxidation of fatty acids.” Experiments were performed to correlate the urinary content of ketone bodies to that of blood and liver, and to the liver energy sources. Sedeficient and Se-supplemented (15 or 20 weeks of supplementation) rats were used, and starved for 48 hours. Se-deficient groups had about 7% lower body weight (434 f 17 g) and the Se-dependent GSH-Px activity was 4% of Se-supplemented controls (Table 1). Glucose content of liver and blood was comparable for both groups and about 9 pmol/g liver and 3.5 pmol/ mL blood, respectively. The glycogen content was about 25 pmol/g liver for both groups, which was about 10% of that of fed rats (not shown). The liver triglyceride content of Se-deficient rats was half that of the supplemented groups (Table 1). However, this difference was apparently not the cause of increased urinary excretion of ketone bodies (Table l), as neither liver nor blood ketone body content was increased in the Se-deficient group (not shown). The AcAc content of liver and blood was about 0.1 hmol per gram and milliliter, respectively, and the value for 3-OHBA was about 0.95 and 0.58 pmol per gram of liver and milliliter of blood, respectively. The urinary amount of AcAc in Fig 2 was lower than that in Fig 1 and Table 1, when expressed per 24 hours urine volume. This difference emerged mainly from a lower urinary volume of rats in Fig 2 than in Table 1 and Fig 1 and not from differences in the per milliliter values. These experiments also indicated that the Sedeficient group did not mobilize liver triglyceride during starvation while the Se-supplemented and standard diet group consumed about 30% (not shown). Although not significant, this was opposite from what could be expected if the urinary AcAc was a result of increased ketone body production in the Se-deficient liver. The increase of liver ketone bodies (AcAc and 3-OHBA) on starvation was 9.5 times for the Sedeficient group and 5.5 times for Se-supplemented group. This difference was not significant and is probably not large enough to explain the enhanced ketonuria in the Se-deficient rats. Preliminary Determination in the Kidney

of Ketone Bodies

One experiment was performed with two Sedeficient and two supplemented rats, to determine the

ketone body concentration in the kidney after 48 hours of starvation. Although only two rats per group were available, a significantly (0.01 < p < 0.02) higher sum of ketone bodies (AcAc + 3-OHBA) was found in the Se-deficient rat kidney, than in the Se-supplemented controls, (means f SE, 0.58 f 0.005 and 0.44 + 0.015 hmol per gram, respectively). This would indicate that the high ketonuria of Se-deficient rats might involve a higher ketone body content in the kidney. However, this difference is probably not the sole cause of the highly significant difference in urinary AcAc excretion between Se-deficient and supplemented rats. More probable is that this difference emerged from increased ketone body production and decreased kidney reabsorption in the Se-deficient rat. DISCUSSION

Unstarved Se-deficient male rats excreted small but significant amounts of AcAc in the urine compared to Se-supplemented controls. Starvation for 24 or 48 hours greatly enhanced this Se-dependent difference. As mentioned above, Se-supplemented rats did not differ from rats fed the standard diet. This should be taken as a further indication that the ketonuria under discussion is a true effect of selenium deficiency. Neither the content of ketone bodies and glucose in blood and liver, nor liver glycogen content was affected by selenium deficiency. Therefore, ketonuria in the Se-deficient rat was not simply caused by increased liver synthesis and ketonemia, as is otherwise the well known sequence on starvation and uncontrolled diabetes.” Other possible causes of this finding might involve a change of ketone body turnover in the Sedeficient rat. To clarify the reasons for this assumption some comments on the ketone body production and utilization can be made. The obligate substrate for ketogenesis is acetylCoA, which accumulates from fatty acid P-oxidation or from pyruvate oxidation in the absence of entry into the citric acid cycle.** Accumulation of acetyl-CoA leads to ketone body formation via the hydroxymethylglutaryl(HMG)-CoA cycle. The rate-limiting enzyme in this cycle is HMGCoA synthase and this enzyme is present in large quantity only in the liver, which marks this tissue as the primary site of ketogenesis. In contrast, the ratelimiting enzyme of ketone body utilization (3-oxoacidCoA-transferase) is present in significant quantities in all tissues except the liver.** However, the ketone body synthesizing enzymes are present in peripheral tissues, and the ketone body utilization enzyme system is reversible21*23and thus not excluded from performing ketogenesis. Fatty acids and ketone bodies are the main fuels of respiration in rat kidney cortex. Ketone body synthesis by kidney may occur through reversal

IMPAIRED

997

KETONE BODY METABOLISM

of the 3-oxoacid-CoA-transferase reactionz4 or by a 25 The finding of enhanced ketonudirect deacylation. ria in the Se-deficient rat could therefore be an effect on the kidney, explainable as decreased utilization or induced synthesis of ketone bodies, or both. So far, a changed turnover of AcAc and 3-OHBA seems to be the most probable explanation, although changes in glucose turnover cannot be entirely excluded.2’ Similarities in ketone body metabolism between most peripheral tissues*’ suggest that selenium deficiency could lead to impaired ketone body metabolism also in other peripheral tissues, such as the heart. The urinary excretion of ketone bodies decreased to about 50% after two days of selenium supplementation (Fig 1). This rapid reversal might indicate an enzymatic role for selenium in ketone body metabolism. Selenium has apparently some role in ketone body metabolism and this is interesting as such. However, there may be a further importance of this finding, with regard to suggested selenium protective roles in atherosclerosis6 and carcinogenesis4”,’ and with regard to evidence for an involvement of radical processes in atherosclerosis26 and carcinogenesis.*’ The reason for a possible connection between selenium, ketone body impairment, and these diseases is

that four mammalian enzymes were reported to catalyze the peroxidation of AcAc through a radical peroxidation process.28 Two of these enzymes were prostaglandin synthetase and prostacyclin synthetase. In this context it is worthy of note that the Se-dependent enzyme glutathione peroxidase might interact with prostaglandin synthesis.* Part of the present finding might thus depend on a change of the prostaglandingoverned glomerular filtration and reabsorption in the Se-deficient kidney. Work is now in progress to find the enzymatic and physiological role(s) for selenium in ketone body metabolism of liver and peripheral tissues. Some possible connections between selenium deficiency, ketone body metabolism, and different diseases have been suggested with the purpose to stimulate specialists in these different fields of research to judge if they can make use of the present finding.

ACKNOWLEDGMENT Thanks are due to Lilly Johansson for skilful technical assistance and to professor Barbara Cannon for help with language and for discussions. Thanks are also due to professor Erik Arrhenius for arousing my interest in selenium research.

REFERENCES 1. Schwarz K, Foltz CM: Selenium as an integral part of “factor 3” against dietary necrotic liver degeneration. J Am Chem Sot 1913292-3293, 1957 2. FlohC L, Giinzler WA, Loschen G: The Glutathione peroxidase reaction: A key to understand the selenium requirement of mammals, in Kharash N (ed): Trace Metals in Health and Disease. New York, Raven Press, 1979, pp 263-286 3. Tappel AL: Measurement of and protection from in vivo lipid peroxidation. in Pryor WA (ed): Free radicals in biology, ~014. New York, Academic Press, 1980, pp 147 4. Shamberger RJ: Biochemistry of selenium. New York, Plenum Press, 1983, pp a) 3 l-58, b) 207-271, c) 227-229 5. Griffin

AC:

Role

of selenium

in the

chemoprevention

of

cancer, in Klein G, Weinhouse S (eds): Advances in Cancer Research, vol 29. New York, Academic Press, 1979. pp 419-442 6. Shamberger RJ: Selenium in health and disease, in Industrial Health Foundation, Inc (ed): Proceedings of the symposium on Selenium-Tellurium in the environment. Pittsburgh, 1976, pp 253261 7. Chen X, Yang G. Chen J, et al: Studies on the relations of selenium and Keshan Disease. Biol Trace Element Res 2:91-107, 1980 8. Burk RF: Biological activity of selenium. Ann Rev Nutr 3:53-70, 1983 9. Paglia DE, Valentine WN: Studies on the quantitative and qualitative characterization of erythrocyte glutathione peroxidase. J Lab Clin Med 70: 158-l 69, 1967 10. Gtlnzler WA, Kremers H, FlohC L: An improved coupled test procedure for glutathione peroxidase (EC I, 11.1.9)in blood. Biothem 12:444-448, 1974 1 I. Lowry OH, Rosebrough NJ, Farr AL et al: Protein measurement with the folin phenol reagent. J Biol Chem 193:265-275, 1951

12. Eggstein M, Kuhlmann E: Triglycerides and glycerol, in Bergmeyer HU (ed): Methods of Enzymatic Analysis, ~014. New York, Academic Press, 1974, pp 1825-l 83 1 13. Keppler D, Decker K: Glycogen, in Bergmeyer HU (ed): Methods of Enzymatic Analysis, vol 3. New York, Academic Press. 1974,pp1127-1131 14. Bergmeyer HU, Bernt E. Schmidt F, et al: D-glucose, in Bergmeyer HU (ed): Methods of Enzymatic Analysis, vol 3. New York, Academic Press, 1974, pp 1196-1201 15. Mellanby I. Williamson DH: Acetoacetate. in Bergmeyer HU (ed): Methods of Enzymatic Analysis. vol 4. New York. Academic Press, 1974, pp 1840-l 843 16. Williamson DH, Mellanby J: D-(-)-3-hydroxybutyrate, in Bergmeyer HU (ed): Methods of enzymatic analysis, vol. 4. New York, Academic Press, 1974, pp 1836-1839 17. Williamson DH, Brosnan JT: Concentration of metabolites in animal tissues, in Bergmeyer HU (ed): Methods of Enzymatic Analysis, ~014 New York, Academic Press, 1974. pp 2291-2292 18. Faupel RP, Seitz HJ, Tarnowski W, et al: The problem of tissue sampling from experimental animals with respect to freezing technique, anoxia. stress and narcosis. Arch Biochem Biophys 148:509-522.1972 19. Williamson DH, Lund P, Krebs HA: The redox state of free Nicotinamide-Adenine Dinucleotide in the cytoplasm and mitochondria of rat liver. Biochem J 103:514-527, 1967 20. Harano Y, Kosugi K, Hyosu T, et al: Sensitive and simplified method for the differential determination of serum levels of ketone bodies. Clin Chim Acta 134:327-336, 1983 2 1. Robinson AM, Williamson DH: Physiological roles of ketone bodies as substrates and signals in mammalian tissues. Physiol Rev 60:143-187. 1980 22. McGarry

JD, Foster DW: Regulation

of hepatic

fatty acid

998

oxidation and ketone body production. Ann Rev Biochem 49:395420, 1980 23. Seifter S, Englard S: Energy metabolism, in Arias IM, Popper H, Schachter D, et al (eds): The liver biology and pathology. New York, Raven Press, 1982, pp 2 19-249 24. Weideman MJ, Krebs HA: The fuel of respiration of rat kidney cortex. Biochem J 112: 149-I 66, 1969 25. Brady PS, Scofield RF, Ohgaku S, et al: Pathways of acetoacetates formation in liver and kidney. J Biol Chem 257:92909293,1982

ULF OLSON

26. McCay PB: Physiological significance of lipid peroxidation. Fed Proc 40:173, 1981 27. Demopoulos HB, Pietronigro DD, Flamm ES, et al: The possible role of free radical reactions in carcinogenesis. J Environ Path01 Toxic01 3:273-303, 1980 28. Harrison JE, Saeed FA: Radical acetoacetate oxidation by myeloperoxidase, lactoperoxidase, prostaglandin synthetase and prostacyclin synthetase: Implications for atherosclerosis. Biochem Med 29:149-163, 1983