Studies on the Mechanism of Activation of Adipose Tissue Pyruvate ...

THE JOURNAL OF BIOLO~KXL CHEMISTRY Vol. 248, No. 1,Issue of January 10,~~. 73-81, 1973 Prnted

in

U.S.A.

Studies on the Mechanism of Activation Pyruvate Dehydrogenase by Insulin*

of Adipose

Tissue

(Received for publication, August 28, 1972) SIMEON

I.

TAYLOR,~

From the Department

L.

CHHABIRANI

MUKHERJEE,AND

ROBERT

of Biological

Chemistry,

Medical School, Boston, Massachusetts

Harvard

SUMMARY Incubation of rat epididymal adipose tissue fragments with insulin led to increases of up to 2.4-fold in the activity of pyruvate dehydrogenase subsequently assayed in tissue homogenates. The activation of pyruvate dehydrogenase by insulin was greatest when tissue was incubated in the presence of bicarbonate ions and when 2 mg of glucose or fructose per ml was added to the incubation medium. Several other agents known to inhibit lipolysis and to decrease cyclic AMP levels in fat cells, including niacin, 5-methylpyrazole-3-carboxylic acid, and prostaglandin El were also effective in activating pyruvate dehydrogenase. Severe depletion of tissue ATP levels caused by the addition of oligomycin or dinitrophenol, by anaerobic incubation, or by prolonged incubation with epinephrine in the absence of albumin also activated pyruvate dehydrogenase. Incubation of adipose tissue with 1 rnM oleate, 1.5 mM octanoate, 1.5 rnx heptanoate, 1 mM butyrate, or 3 lflM DL-P-hydroxybutyrate decreased pyruvate dehydrogenase activity 20 to 70%. The addition of 5 my acetate, 5 mM propionate, or 3 rnM 4-pentenoate led to 70 to 160% activation of pyruvate dehydrogenase, whereas 1.5 mM pentanoate had no effect. High concentrations of glucose (20 mg per ml) or of pyruvate-lactate (5 and 30 mM, respectively) also increased tissue pyruvate dehydrogenase activity.

The well known ability of insulin to accelerate fat.ty acid synthesis from glucose in adipose tissue (1) was considered for many years to bc a simple consequence of the facilitation of glucose transport by this hormone (2). The development of the tritiated water technique for measuring rates of lipogenesis without resort to %-labeled precursors (3,4) made possible the study of lipogenesis from endogenous substrate. Such studies demonstrated unequivocally that insulin activates one or more steps in * This research was supported by Public Hea1t.h Service Research Grant AM-08076 from the National Institute of Brthritis and Metabolic Diseases. $ Supported by the Insurance Medical Scientist Scholarship Fund from the Massachusetts Mutual Life Insurance Co., Springfield, amass. Q Recipient of Public Health Service Research Career Development Award AM-10556.

73

JUNGAS~

02115

the lipogenic pathway subsequent to glucose entry (5). It was possible to identify one such step as the reaction catalyzed by pyruvate dehydrogenase (5-9). Fortunately, at about the same time, studies in Reed’s laboratory of the enzymology of the pyruvatc dchydrogenase complex of mammalian heart, kidney, and liver led to the proposal of a model describing the regulation of the activity of this enzyme (10, 11). According to this model, pyruvatc dehydrogenase is regulated by a phosphorylation-dephosphorylation mechanism; the dephosphorylated enzyme is the active form, whereas the phosphorylatcd enzyme is totally inactive. Several laboratories have obtained evidence that this model applies to pyruvate dehydrogenase of rat adipose tissue (6, 8, 9). Thus it has been proposed that the activation of adipose tissue pyrnvate dehydrogenase by insulin is associated with a decreased degree of phosphorylation of this mitochondrial enzyme (5-9). We have now extended our studies of the mechanism of activation of pyruvate dehydrogenase by insulin in adipose tissue. The action of insulin is compared to that of several other antilipolytic agents capable of decreasing cellular levels of cyclic AMP. The influence of bicarbonate ions and of a variety of oxidizable substrates on pyruvate dehydrogenase activity is described. Preliminary reports of some of these results have appeared previously (12, 13). MATERIALS

AiVD

METHODS

Rats were obtained from the Charles River Breeding Laboratories, Inc., North Wilmington, Mass., and were maintained on Purina Laboratory Chow ad lib&m for 1 to 4 weeks before their use. Rats in the weight range 150 to 300 g were selected for most experiments. In experiments in which tissue from fasted-refed animals was employed, the animals were fasted for 3 days and then given free access for 2 days to the high carbohydrate, fat-free diet devised by Dr. J. P. Flatt (5), obtained in pelleted form from General Biochemicals, Inc. Rats were killed by decapitation, and the epididymal fat bodies were removed and floated in 0.15 M NaCl so that thick portions could be dissected out and discarded. The remaining thin segments were divided into an appropriate number of pieces and distributed among the incubation flasks such that control and experimental values were always obtained on tissues from the same animal(s). Statistical tests of significance used Student’s t test for paired or unpaired data as appropriate for each case. Krebs-Ringer bicarbonate (14) and Krebs-Ringer phosphate (15) media were prepared with one-half of the recommended

74 calcium. In incubations with Krebs-Ringer bicarbonate the gas phase was 5:1’, CO2 in air, and with Krebs-Ringer phosphate the gas phase was air. 0 1zl inarily, tissue fragments (100 to 300 mg) mere incubated in 2 ml of appropriate medium in a lo-ml Erlenmcyer flask for 30 min at 37” before transfer to 2 ml of fresh, prewarmed, prrgnssed medium for the experimental incubation period. To assay whole t.issue pyruvate dehydrogenase activity, the adipose tissue (100 to 300 mg) was homogenized in a Ten Broeck homogrnizer containing 0.4 ml of ice-cold buffer (10 mM potassium phosphate, 1 IBM EI)TA, I mM dithiothreitol, 1% bovine serum albumin, pH 7.4). Aliquots of 0.2 ml were immediately transferred to the main chamber of lo-ml Erlenmeyer flasks cont.aining 0.3 ml of an assay mixture (see below) and warmed to 37”. The flasks possessed center wells fitted with a gelatin capsule containing a strip of glass fiber paper (1 cm x 2 cm) and The flasks were closed with a rubber 0.2 ml of phenethylamine. serum stopper a.nd allowed to incubate for 2 min at 37” with shaking. The composition of the assay mixture was adjusted to give the following final concentrations during the assay: 11 mM potassium phosphate, 1.1 mM EDTA, 2.8 mM MgC12, 1.6% serum albumin, 1.2 1llM dithiothreitol, 0.16 InM CoA, 1.6 nm NAT)+, 0.08 111~ thiamine pyrophosphate, 0.6 mM pyruvate, 0.12 PCi of [l-14C]pyruvate per ml (pH 7.4). The reaction was stopped by injecting 0.8 ml of 0.08 M citric acid 0.04 M Na2HP04 buffer (pH 3). After 30 min of further incubation the gelatin capsule and its contents were removed by forceps and placed in a 3-dram vial containing 5 ml of a toluene-ethanol (9: 1, v/v) scintillation fluid (0.4% Omnifluor, New England Nuclear). The 14C was measured in either an Ansitron or a Packard scintillation counter. Results were corrected for blank values obtained in vessels in which 0.2 ml of homogenizing buffer replaced the tissue homogenate. Such blank values were typically in the range of 600 to 1200 cpm, depending on the lot and age of the [I-l%]pyruvate sample employed. Experimental values were 3 to 10 times the blank values. Several precautions were taken to minimize the blank corrections. Upon receipt the sodium [I-‘4CJpyruvate was distributed into small vials of 5 PCi each, lyophilized, and stored dry at -75” under air. For use, 50 11 of water was added to a vial, and an aliquot appropriate for the experiment at hand The vial was immediately removed and diluted in 0.004 N HCl. was then immediately refrozen at -75” and stored for periods up to 4 days. The acidified solution of sodium [1-‘4Clpyruvate was kept on ice and added to the assay mixture, also on ice, just before use. The assay vessels charged with the assay mixture were kept on ice until they were placed in the 37” water bath for 1 min before addition of the tissue extract. Thin layer chromatography of the labeled pyruvate samples did not prove useful. Glycogen phosphorylase activity was assayed as described earlier (16), and t’he tissue content of free fatty acids was assayed by the method of Dole and Meinertz (17). Tissue content of ATP was assayed by a modification of the method of Halperin and Denton (18). The freezing of the tissue in liquid nitrogen was found to be unnecessary and was omitted. The tissue (approximately 150 mg) was homogenized in a Ten Broeck homogenizer containing 5% perchloric acid (0.5 ml) and ether (1.0 ml) at 0”. The homogenizer was washed with additional 5% perchloric acid (0.5 ml) and ether (1.0 ml). After removing the lower aqueous phase from the combined extract and washings, the organic phase was washed with 100 mM triethanolamine hydrochloride (0.1 ml) and 5% perchloric acid

(0.2 ml). The aqueous phases were combined and neutralized with 10 x KOII (0.05 ml) and 2 M KzHPOd (0.2 ml). Following centrifugat,ion, the supernatant was decanted and subjected to a water pump vacuum for 10 min. ATP was assayed in the resulting extract by the use of hexokinase and glucose 6-phosphate dehydrogenase, measuring appearance of NADPH with a Turner fluorometer (model 110). [l-14C]Pyruvate oxidation by intact tissue was measured by collecting 14C02 in plastic disposable center wells (Kontes Glass) containing fiber glass paper strips and phenethylamine. The incubation was terminated by acidifying the incubation medium with 1.25 volumes of 0.08 M citrate-O.04 M phosphate buffer (pH 3), and CO2 was collected and counted as described for the nyruvate dehydrogenase assay. For t.he assay of pyruvate in incubation media, 2.5.ml aliquots were acidified with 0.6 ml of 6 N HzSO.+ The acidified media were stored overnight at 4” and then neutralized with 0.35 ml of 10 x KOH and 0.5 ml of 2 M KzHPO4. Pyruvate was assayed spectrophotometrically by using lactate dehydrogenase. Insulin, epinephrine bitartrate, CoA, NAD+, and sodium pyruvate were obtained from Sigma Chemical Corp. [I-r%]Pyruvate, sodium salt, was purchased from New England Nuclear Corp. Niacin, propionic acid, pentanoic acid, heptanoic acid, and octanoic acid were obtained from Eastman Chemical Corp. Thiamine pyrophosphate, calcium lactate, and sodium m-P-hydroxybutyrate were obtained from Calbiochem. 4-Pentenoic acid was obtained from K & K Laboratories, Plainview, N. Y. Butyric acid and phenethylamine were purchased from Fisher Chemical Co. Bovine serum albumin (Fraction V, fatty acid poor) was obtained from Miles Laboratory, Kankakee, 111. Glucose g-phosphate and lactate dehydrogenase, hexokinase, and NADf were supplied by Boehringer Mannheim Corp., New York. Gelatin capsules size 0, were obtained from Eli Lilly & Co. PGEI1 and 5-methylpyrazole-3-carboxylic acid were the generous gifts of Dr. John E. Pike and Dr. George C. Gerritsen, respectively, of the Upjohn Company, Kalamazoo, Mich. RESULTS

E$ects of Bicarbonate Ions and Hexoses on Respme to InsulinWe have previously reported that incubation of adipose tissue fragments with insulin leads to small increases in the pyruvate dehydrogenase act.ivity observed in whole homogenates of the tissue (5, 6). The medium used for the incubation of the tissue in these experiments was Krebs-Ringer phosphate containing no added substrate. Subsequently, other workers employing Krebs-Ringer bicarbonate medium often supplemented with fructose or glucose have reported considerably larger effects of insulin (7-9). We have therefore investigated the influence of changes in the incubation medium on the magnitude of the insulin effect. It was found that the activation of pyruvate dehydrogenase by insulin was considerably larger when the tissue was incubated in Krebs-Ringer bicarbonate medium (Fig. 1). In particular it was the presence of bicarbonate which was required rather than the absence of large amounts of phosphate. Furthermore, the effect of insulin is greatly enhanced when fructose or glucose is included in the incubation medium (Fig. 2). We emphasize that there is a consistent effect of insulin on pyruvate dehydrogenase activity, albeit a small one, even in the absence of added hexose. Although insulin augments the utilization of glucose by adipose tissue to a greater extent than the utilization of fructose (8, 19), the effect of the hormone on pyr1 The abbreviations used are: PGEI, prostaglandin AMP, cyclic adenosine 3’,5’-monophosphate.

E,; cyclic

75

d Y---v

‘0 t > $ ‘L 0

60

-

40-

k---$

OC

Niacin

+---45-MP-3-CA

20’

51

.

-

1 9 0

I 1 * I I L Id’0 10-g lo-* lo-7 lo-6 105 10-4

I

Concentration-M

05

HC03

PO4

ttco3+

PO4

FIG. 1. Requirement for bicarbonate ions in the activation of pyruvate dehydrogenase (PDH) by insulin. Tissues were incubated for 30 min in 2 ml of either Krebs-Ringer bicarbonate (HCO,), Krebs-Ringer phosphate (POl), or bicarbonate-phosphate (HC03 + PO*) medium containing 2 mg of fructose per ml. The bicarbonate-phosphate medium consisted of Krebs-Ringer bicarbonate with the sodium, chloride, and ph0sphat.e concentrations modified to 128, 98, and 16.2 mM, respectively. The tissues were then transferred to fresh media of the same composition, with or without insulin (10 milliunits per ml) as indicated. After 30 min, the tissues were homogenized and assayed for pyruvate dehydrogenase activity. Ppruvate dehydrogenase activities, expressed as micromoles per g of fresh tissue in 10 min, are the averages of eight observations, with the S.E. given by vertical lines. The insulin effect was significant in both bicarbonate-containing media (p < 0.001) but not in Krebs-Ringer phosphate (p = 0.11).

-1

Fructose

No Fructose

Fructose

Glucose

FIG. 2. Substrate requirements for maximal effect of insulin on pyruvate dehydrogenase (PDH). Tissue fragments were incubated in Krebs-Ringer bicarbonate by using the double incubation procedure described in Fig. 1. Hexoses (2 mg per ml) were present during both incubations, whereas insulin (10 milliunits per ml) was added only during the second period. Results are the averages of six observations, with the S.E. given by the vertical lines. In order to minimize variations among animals, the data have been normalized by setting the average control activity in the presence of fructose for each experiment equal to 1. The observed mean control activities for the group of experiments summarized wa.s 0.77 pmoles per g in the Eeft-hand panel and 0.82 pmoles per g for the right-hand panel, in 10 minutes. The insulin effect, was not significant in the absence of hexose (p = 0.08).

FIG. 3. Dose-response curves for the activat.ion of pyruvate dehydrogenase by niacin and 5-methylpyraxole-3-carboxylic acid (~-MPG-CA). Tissues were incubated in Krebs-Ringer bicarbonate containing fructose (2 mg per ml) by using the double incubation procedure described in Fig. 1. Niacin and 5-MP-S-CA were present only during the second period. Data are the average f S.E. of six observations. Kinetic parameters K, the affinity constant, and PO, the maximal response, were estimated by a least squares fit to a hyperbola of the form P = 1 + PO-D/(D + K), where D is the drug concentration (30). Results for niacin (K = 0.05 f 0.03 PM, PO = 1.11 f 0.11) and 5-MP-b-CA (K = 0.06 i 0.03 PM, PO = 0.35 f 0.03) were obtained on different rat.s and are expressed as a percentage of PO to facilitate comparisons. Activations of pyruvate dehydrogenase by both agents were significant at concentrations of 0.08, 0.8, and 8 PM (p < 0.05).

uvate dehydrogenase appears to be enhanced equally by either hexose. E$ects of Antilipolytic Agents-Insulin is known to inhibit lipolysis (20, 21) and to decrease the cyclic AMP content of adipose tissue under some circumstances (22, 23). Therefore, it was of interest to determine whether other agents which inhibit lipolysis and decrease cyclic AMP levels share insulin’s ability to activate pyruvate dehydrogenase. We have examined three such agents: niacin (24, 25), 5-methylpyrazole-3-carboxylic acid (26, 27), and PGE, (28, 29). Previous studies utilizing tissue from normally fed rats demonstrated that niacin (80 pM) did cause an increase in pyruvate dehydrogenase activity and that this effect also was enhanced when bicarbonate ions were present (13). Fig. 3 shows the dose-response curves for the effects of niacin and 5-methylpyrazole-3-carboxylic acid upon pyruvate dehydrogenase activity of adipose tissue from normally fed rats. These agents caused half-maximal activation of pyruvate dehydrogenase at concentrations of 0.05 f 0.03 pM (niacin) and acid). In 0.06 f 0.03 PM (5-methylpyrazole-3-carboxylic separate experiments with adipose tissue from fasted-refed rats, niacin caused a half-maximal inhibition of glycerol production at Uutcher has reported halfa concentration of 0.08 f 0.13 PM. maximal decreases in cyclic AMP content of adipocytes at concentrations of niacin and 5-methylpyrazole-3-carboxylic acid of approximately 0.08 PM (27). Thus these data are consistent with a role for cyclic AMP as an intracellular mediator of the actions of niacin and 5-methylpyrazole-3-carboxylic acid on both glycerol release and pyruvate dchgdrogenase activation. We have shown previously that niacin and insulin both activate pyruvate dehydrogenase in adipose tissue from fastedrefed rats (12, 13). However, as shown in Fig. 4, even high concentrations of niacin failed to influence the glycogen phosphorylase activity of adipose tissue from fasted rats. The phosphorylase activity in such tissue is pa.rticularly sensitive to the action of insulin, an effect thought to be mediated by cyclic

76

r

r

Control PGE,

L

Control

Ins

Nio

Control

Epi

Epi Nio

FIG. 4. Effects of niacin (Nia) on glycogen phosphorylase activity of adipose tissue from fasted-refed and normal rats. Tissue from fasted-refed rats (A) was incubated for 1 hour in 2 ml of Krebs-Ringer bicarbonate medium containing insulin (Ins; 10 Phosphorylase milliunits per ml) or niacin (80 PM) as indicated. activity was assayed in whole tissue homogenates without added AMP. Data are expressed as micromoles of phosphate per 100 mg of fresh tissue in 15 min and are the averages and S.E. of six experiments. The inhibition of phosphorylase by insulin was significant (p = 0.05), but niacin had no effect (p = 0.84). Tissue from normally fed rats (B) was incubated for 40 min in 2 ml of Krebs-Ringer bicarbonate medium containing fructose (2 mg per ml). Niacin (80 PM) was then added; 10 min later epinephrine After 10 min, tissue was (Qi; 2 &M) was added as indicated. assayed for phosphorylase without added AMP. Results, expressed as in A, are averages and S.E. of six experiments. Activation by epinephrine was significant (p = 0.00005) and was partially blocked by niacin (p = 0.01). AMP (16). In tissue from normally fed rats, niacin does diminish the activation of phosphorylase brought about by epinephrine (Fig. 4), an effect consistent with its ability to lower cyclic AMP in this tissue. A substance of very different chemical structure, PGEi, is also able to inhibit lipolysis and to decrease cyclic AMP levels in adipocytes. As illustrated in Fig. 5, PGEi in concentrations of either 2.8 or 28 PM activates pyruvate dehydrogenase in tissue of normally fed rats. In these experiments, 1% ethanol was present in all incubation media since PGEi was added from an ethanolic stock solution. By itself, ethanol at this concentration increased pyruvate dehydrogenase activity only slightly. Pyruvate Dehydrogenase Activity and Tissue Content of ATPBecause pyruvate dehydrogenase in a variety of tissues can be inactivated by an ATP-dependent kinase (6-11, 31), it seemed possible that agents which increased pyruvate dehydrogenase activity might do so by depleting tissue ATP levels. Since pyruvate dehydrogenase is a mitochondrial enzyme (32), it is presumably the mitochondrial pool of ATP which serves as substrate for the pyruvate dehydrogenase kinase. Therefore the effects of agents which interfere with the synthesis of ATP by mitochondrial oxidative phosphorylation were examined. Data in Table I, Experiment 1, demonstrates that four agents capable of depleting the tissue content of ATP also activated pyruvate dehydrogenase: dinitrophenol, an uncoupler of oxidative phosphorylation (33); oligomycin, an inhibitor of energy transfer reactions associated with oxidative phosphoryIation (34); anaerobiosis; and prolonged incubation with large amounts of epinephrine, which leads to massive accumulations of fatty acids

2.8pM

PGE, 28pM

Niacin 8,,M

FIG. 5. Activation of pyruvate dehydrogenase (PDH) in normal adipose tissue by PGE, and niacin. Tissues were incubated as described in Fig. 3. Additions during the second incubation period were: fructose (2 mg per ml), ethanol (lyO), and PGE, and niacin as indicated. Results were normalized as described in Fig. 2 and are averages and S.E. of 17 (PGEI, 28~~), 5 (PG&, 2.8 gM), or 8 observations (niacin, 8 pM). The observed average control activity was 1.81 rmoles per g in 10 min. Activations caused by 2.8 PM PGE,, 28 PM PGE,, and 8 PM niacin were significant, with p = 0.05, 0.0005, and 0.01, respectively. in t.he tissue which in turn uncouple and inhibit the respiratory chain (35, 36) and thus decrease tissue ATP (37, 38). This effect of epinephrine to activate pyruvate dehydrogenase can be prevenOed by including glucose and albumin in the incubation medium (13), presumably because the accumulation of tissue fatty acids is thereby greatly diminished. However, neither insulin nor niacin decrea.ses tissue ATP levels (Table I, Experiment Z), and thus their effects on pyruvate dehydrogenase can not be accounted for on this basis. When insulin is added to the incubation medium along with the epinephrine, the fall in tissue ATP is nearly obliterated, whereas pyruvate dehydrogenase activity is maintained at a high level (Table I, Experiment 3). This suggests that insulin may activate pyruvate dehydrogenase even in the presence of epinephrine. This interpretation is supported by the observation that the addition of pyruvate, which is less effective in preventing the decline in tissue ATP levels caused by epinephrine, nevertheless allows the pyruvate dehydrogenase activity to fall below the level found with insulin and epinephrine. E$ects of Substrates upon Pyruvate Dehydrogenase ActzvityData summarized in Table II show that very high concentrations of glucose in the incubation medium activate pyruvate dehydrogenase. Even at the highest concentration of sugar tested, insulin was still able to produce a further activation of pyruvate dehydrogenase. Similar results have been obtained with fructose (13). These results may be contrasted with the effects of pyruvate (Fig. 6). In these experiments a 6-fold excess of lactate was added to the incubation medium along with pyruvate in order to minimize changes in the cytoplasmic oxidation-reduction potential. The addition of lactate by itself at these concentrations had little influence on pyruvate dehydrogenase activity (data not presented). Whereas pyruvate-lactate increased pyruvate dehydrogenase in a manner similar to that of glucose or fructose, at high concentrations of pyruvate-lactate no additional effect of insulin could be demonstrated.

77 TABLE

Effects

of

1

various substances on pvruvate dehydrogenase Al’P

content

of adipose

activity and

tissue

In Experiments 1 and 3, tissue fragments were incubated for 80 min in Krebs-Ringer phosphate medium without albumin or glucose. Additions to the medium were present from the start at the concentrations indicated. The anaerobic vessel in Experiment 1 was gassed with nitrogen during the first 15 min of the incubation period. In Experiment 2, tissues were incubated for 75 min in Krebs-Ringer bicarbonate containing 1 mg of glucose per ml with other additions present from the start. In Experiment 4, the tissues were incubated for 30 min in Krebs-Ringer bicarbonate with 2 mg of fructose per ml. They were then transferred to vessels containing the same buffer supplemented with 2 mg of fructose per ml, 1.25% serum albumin, and the additions indicated for a further 45-min incubation period. Values are the means and standard errors for the number of experiments indicated in parentheses. Pyruvate Senase

Additions

dehydroactivity

TABLE II upon pyruvate dehydrogenase activity Tissues were incubated for 30 min in Krebs-Ringer bicarbonate containing glucose, 1 or 20 mg per ml, before transfer to identical media, with or without insulin, for an additional 30-min incubation. Results are expressed as the average G.E. of three experimerits . The effect of insulin was significant at both low (p = 0.0007) and high (p = 0.007) glucose concentrations; the glucose effect was significant only in the absence of insulin (p = 0.01).

E$ects

Addition

0.86 2.95 2.57 2.79 3.10

zk f zk f f

0.15 0.83 0.44 0.30 0.66

Experiment 2 Control.. . . Insulin, 10 milliunits/ml . Niacin, 80 PM.. . . .. Experiment 3 Control.. . . .. . Epinephrine, 2 j&M. . . Epinephrine and insulin, 1 milliunit/ml . Epinephrine and pyruvate, 10rnM ._................ Experiment 4 Control.. . .. Insulin, 1 milliunit/ml. Acetate, 5 mM.. . Butyrate, 1 MM.. .

. ..

Glucose,

Control. . . . . . . .. . Insulin, 10 milliunits/ml

115 f 41 f 22 3~ 24f6 20 i

13 (5) 4 (5) 6 (5) (5) 7 (5)

190 f 200 f 200 f

60 (3) 30 (3) 20 (3)

= m

‘I

Glucose,

1.78 & 0.11 3.33 & 0.11

.

20 n&ml

2.81 i 0.16 3.73 zlc 0.14

Control Insulin

5 I

i l/6

None

f

O.OG

3.10 f

0.22

114 f 34 f

21 (5) 3 (5)

2.29 f

0.11

92 f

14 (5)

1.61 f

0.07

78 zk 6

0.90

1 m&ml

ATP content

2 Experiment 1 Control.. . . . .. Epinephrine, 2 PM. Dinitrophenol, 100 jtM.. . . Oligomycin, 10 ag/ml . Anaerobic . .

of glucose

210 210 200 180

* 40 zk 30 * 40 Z!Z60

(5) (3) (3) (3) (3)

The ability of antilipolytic agents to activate pyruvate dehydrogcnase suggested the possibility that free fatty acids might play a role in modulating the activity of this enzyme. Fig. 7 illustrates the decrease of pyruvste dehydrogenase activity Khich resulted when oleate was added to the incubation medium. rn&Hydroxybutyrate resembled oleate in this respect (Fig. 7), suggesting that the inactivation of pyruvate dehydrogenase may be related in some way to the metabolism of acetyl-CoA. At the concentrations used, neither oleate nor DL-P-hydroxybutyrate interfered with the activating influence of large concentrations of insulin. We next tested a series of carboxylic acids containing straight hydrocarbon chains from 2 to 8 carbon atoms in length (Fig. 8). Butyrate, heptanoate, and octanoate, like oleate, were inhibitory. However, pentanoate had no effect, whereas the addition of or 4-pentenoate produced increases in acetate, propionate, Neither acetate nor butyrate pyruvate dehydrogenase activity. altered tissue levels of ATP (Table I, Experiment 4). Butyrate

Pyruvote

/Lactate

5/30 Concentrations

mM FIG. 6. Effects of a pyruvate-lactate mixture on pyruvate dehydrogenase (PDH) activity. The double incubation procedure described in Fig. 1, using Krebs-Ringer bicarbonate medium was employed. Insulin (10 milliunits per ml) was present only during the second incubation period, whereas substrates were present during both periods. The designations 1/6 and 5/30 mean a pyruvate concentration of 1 and 5 mM and a lactate concentration of 6 and 30 mM, respectively. Results are expressed as in Fig. 2 and are the averages of six experiments. The observed average cont.rol activity in the absence of substrate was 1.00 pmoles per g in 10 min. The insulin effect was significant in the absence of substrate (p = 0.006), or at the lower concentrations of substrate (p = O.Ol), but not at the higher substrate concentrations. The activations caused by pyruvate-lactate at concentrations of l/6 and 5/30 were statistically significant (p < 0.0001).

interfered with the ability of insulin to activate pyruvate dehydrogenase (Fig. 9), whereas acetate activated pyruvate dehydrogenase even in the presence of insulin. In Table III we compared the effect of octanoate upon pyruvate dehydrogenase activity as assayed in homogenates, with its effects on the oxidation of [PC]pyruvate and pyruvate uptake in tissue fragments. The inhibition of pyruvate dehydrogenase activity was more marked than the inhibition of either pyruvate uptake or pyruvate oxidation and thus is sufficient to account for the decrease in pyruvate utilization by the intact tissue. This is in contrast to the situation with rat heart where the addition of palmitate to the perfusion medium diminished pyruvate utilization to a greater extent than it decreased homogenate pyruvate dehydrogenase activity (39), suggesting that additional inhibitory mechanisms were involved.

78

Control

Ins

Oleate

ins

Ins + Oleate

P-OHBut

Ins goiLI”t

FIG. 7. Effects of o1eat.e and P-hydroxybutyrate on pyruvate dehydrogenase (PDH) activit#y in the presence and absence of insulin (Ins). Tissues were incubated as described in Fig. 3. Additions during the second incubation period were: albumin (1.25’%), fructose (2 mg per ml), and where indicated 1 rnM oleate, 3 mM DL-@-hydroxybutyrate, and insulin (1 milliunit per ml). Pyruvate dehydrogenase activities are averages of 12 observations expressed as in Fig. 2. The true average control values were 1.79 and 1.00 pmoles per g in 10 min for the left and right panels, respectively. Activations caused by insulin and inhibitions caused by oleate or nL-/3-hydroxybutyrate in the absence of insulin were all statistically significant (p < 0.02).

Control

C2

Insulin

( I mu/m

c3

c4

c5:I

I)

FIG. 9. Effects of carboxylic acids on adipose tissue pyruvate dehydrogenase (PDH) activity in the presence of insulin. Tissues were incubated exactly as described in Fig. 8, except that insulin (1 milliunit per ml) was present in all media during the second incubation period. Results are the averages of eight observations with the S.E. given by vertical lines. Data have been normalized by setting t,he average activity in the presence of insulin and fructose only equal to 2.00. The observed average activity for t.he experiments with acetat.e was 1.32, for propionate, 2.65, for butyrate, 2.67, and for 4-pentenoate, 1.91 pmoles per g in 10 min. The effects of acetate (p = 0.03) and butyrate (p = 0.0002) were significant, but those of propi0nat.e (p = 0.12) or 4-penten0at.e (p = 0.45) were not significant. TABLE

Comparison

of

III

of octanoate on pyruvate

the effects

dehydrogenase activity and pyruvate metabolism Tissue fragments were incubated for 45 min in 2 ml of KrebsRinger bicarbonate containing 2 mM pyruvate. The tissues were then transferred to 25-ml Erlenmeyer flasks containing 4 ml of the above medium supplemented with 1.25’% albumin and with or without octanoate, 1.5 mM, for a second 45-min incubation. To one-half of these flasks, 0.24 &i of [I-Wjpyruvate was added to allow measurement of the rate of pyruvate oxidation, and the remaining flasks were used for the assay of pyruvate dehydrogenase activity and -I pvruvate uptake from the medium. Results are the averages f S.E. of four experiments. Measurement

Control

C2

C3

C4

C5

C7

C8

Control

C5:,

8. Effects of carboxylie acids on pyruvate dehydrogenase activity. Tissues were incubated as described in Fig. 3. Additions during the second incubation period were albumin (1.25%), fructose (2 mg per ml), and where indicated sodium salts of the following acids: acetate CC,), 5 mM; heptanoate (C,), 1.5 mM; octanoate (C,), 1.5 mM; and 4-pentenoate (CS,~), 3 mM. Pyruvate dehydrogenase activities, expressed as in Fig. 2, are averages of eight experiments. The observed average control value in micromoles per g in 10 min for the experiments with acetate was 0.67, for propionate, 1.28, for butyrate, 1.82, for 4-pentenoate, 1.25, and for the others, 1.13. Changes in pyruvate dehydrogenase activity caused by the various acids were all significant (p < 0.02) with t.he exception of pentanoate. FIG. (PDH)

DISCUSSION

The results presented here demonstrate that incubation of rat epididymal adipose tissue fragments with various substrates, hormones, and pharmacological agents alters pyruvate dehy-

pm&s/g

Pyruvate dehydrogenase activity......... Pyruvate oxidation. Pyruvate uptake..

0.97 f 1.25 f 1.35 f

AXWag. inhibition

Octanoate

0.06 0.05 0.08

in 10 mitt

0.37 f 0.95 f 0.99 f

%

0.06 0.05 0.11

63” 23b 26”

a p = 0.002. b p = 0.04. c p = 0.09.

drogenase activity subsequently assayed in whole homogenates water is diluted 30- to lOOof the tissue. Because intracellular fold during homogenization and all cofactors and substrates of pyruvate dehydrogenase are present in excess during assay, our measurements of pyruvate dehydrogenase activity are not likely to be affected directly by changes in the tissue concentration of substrates, inhibitors, or allosteric effecters of pyruvate dehydrogenase. Indeed, the tissue was homogenized in a medium of

79 low osmolarity to insure that the mitochondria would be disrupted, thereby exposing the pyruvate dehydrogena.se as fully as possible to the assay medium. It is necessary therefore to interpret our observations in terms of a model which includes several forms of pyruvate dehydrogenase of differing enzymatic activity, which are only relatively slowly interconvertible, such as the phosphorylation-dephosphorylation model proposed by Linn et al. (10). We envision that the hormones, etc., alter the activity of pyruvate dehydrogenase by altering the degree of phosphorylation of the enzyme protein. Evidence to support this interpretation in the case of insulin has been presented by Coore et al. (8). In the assay employed here the evolution of X!OS from [l‘4CJpyruvate was utilized as a measure of pyruvate dehydrogenase activity in crude homogenates. Kneer and Ball (40) have suggested that an exchange reaction between pyruvate and CO2 can occur in adipose tissue: CH&O%OOH

+ CO2 *

CH&OCOOH

+ r4C02

This reaction presumably involves the reversible rea.ctions catalyzed by the enzymes pyruvate carboxylase, malate dehydrogenase, and fumarase. Under our assay conditions the concentrations of ATP, bicarbonate, acetyl-CoA and NADH are small, and the above exchange reaction should proceed only very slowly. Direct evidence that our assay actually measures pyruvate dehydrogenase activity rather than the above exchange reaction is provided by the observations that the liberation of labeled CO2 requires the addition of NAD+ and CoA, that it is stimulated by thiamine pyrophosphate, and that it is inhibited over 85% by arsenite.2 Furthermore, the addition of an ATP-generating system (i.e. ADP, creatine phosphate, and creatine kinase) to the tissue extracts leads to a time-dependent inhibition of the reaction by greater than 80%.” Whereas ATP shouId inhibit pyruvate dehydrogenase, it would be expected to stimulate the above exchange reaction. Finally, the results obtained with this assay are in substantial agreement with reports from other laboratories which use a different assay (8, 9). In the case of the oxidation of [1-14Clpyruvate by intact tissue (Table III), exchange reactions probably do make important contributions to the results obtained. A major question which arises upon consideration of the data presented concerns the mechanism by which insulin is able to influence the activity of pyruvate dehydrogenase, an enzyme known to be confined within the mitochondrion (32), apparently without itself entering the fat cell (41, 42). Some intracellular component(s) must serve as the mediator of this action of insulin, i.e. there must be an intracellular “messenger.” The intriguing possibility that there is a unique second messenger for insulin which has so far remained undetected and which serves as the intracellular mediator of this and perhaps other actions of insulin cannot be excluded. All substances which decrease cyclic AMP levels in fat cells, which we have tested, also activate pyruvate dehydrogenase. This is true of compounds which differ greatly in chemical structure, namely PGEr, niacin, insulin, and 5 methylpyrazole3carboxylate, and a priori it seems unlikely that these agents would all induce the formation of a unique insulin messenger. Cyclic AMP is further implicat,ed in the regulation pyruvate dehydrogenase activity by the observation that tissue exposed briefly to epinephrine has a lower pyruvate dehydrogenase activity (5,8), and that either adrenocorticotropic hormone or dibutyryl cyclic AMP block the stimulatory effect of 2 R. L. Jungas and S. I. Taylor, unpublished

results.

insulin on pyruvate dehydrogenase activity (8). Moreover, ouabain, an inhibitor of adipocyte adenylate cyclase (43), also produces an elevation in pyruvate dehydrogenase activity, as does incubation in a K+-free medium (44). To date, all attempts in this laboratory to demonstrate a direct influence of cyclic AMP either on pyruvate dehydrogenase itself or on the kinase or phosphatase involved in its regulation (6, 13) have been without success. Similar negative findings have been reported by others (8, 45), and reports that cyclic AMP activates pyruvate dehydrogenase in cell-free systems (46, 47) have not been confirmed. We are led therefore to the view that cyclic AMP exerts its effects on pyruvate dehydrogenase activity indirectly, i.e. that changes in the cytoplasmic cyclic AMP levels induce metabolic alterations which cause changes in mitochondrial metabolite levels, one or more of which acts as a third messenger to modulate the degree of pgruvate dehydrogenase phosphorylation. This hypothesis receives support from the observation that bicarbonate and small amounts of metabolizable substrates are required if the full effect of insulin, and thus, presumably, the effect of decreasing cyclic AMP, on pyruvate dehydrogenase is to be seen. Effects of insulin on lipolysis and glycogenolysis on the other hand are fully expressed in phosphate media with no added substrate (16, 29). Our results with niacin, PGEI, and 5-methylpyrazole-3-carboxylate differ in some respects from those described by Coore et al. (8) and by Martin et al. (44). These workers reported inhibition of glycerol production by each of these three agents at low concentrations which had no influence on pyruvate dehyWe have been unable to find concentrations drogenase activity. of any of these substances which affect glycerol production, and thus presumably fat cell cyclic AMP levels, which do not also lead to activation of pyruvate dehydrogenase. For example, Martin et al. (44) found no activation of pyruvate dehydrogenase by either 1 PM niacin or 10 PM 5-methylpyrazole-3-carboxylate, whereas Fig. 3 shows that concentrations of either substance as low as 0.1 pM were effective in our hands. However, whereas our normal experience has been for the maximal activation of pyruvate dehydrogenase by niacin to resemble the maximal response to insulin, with one batch of rats the maximal response to niacin (38%) and to 5-methylpyrazole-3-carboxylate (35%) was unusually poor and well below that of insulin (103 ~~0). We do not understand the cause for such variable responses between different groups of rats but similar effects may well account for the differences noted between our results and those of Martin et al. (44). The suggestion that the activation of pyruvate dehydrogenase by insulin is quite indirect and involves several intracellular messengers is strengthened by the observat.ion that a small amount of an oxidizable substrate such as glucose, fructose, or pyruvate must be present to obtain the maximal activation. We do not consider it likely that the effect of insulin on pyruvate dehydrogenase activity is secondary to its well known effects on substrate transport. The main reason for this view, discussed more fully elsewhere (13), is that the substrate requirement can be satisfied by low concentrations of pyruvate (1 mM) whose transport is thought to be unaffected by insulin, or by endogenous substrate in the case of tissue from fasted-refed animals. Moreover this requirement for added substrate was not observed by Coore et al. (8), perhaps because of differences in the diets or the strain of rats employed. The ability of large amounts of glucose, fruct.ose, or pyruvate, or of other substrates to cause increases in pyruvate dehydrogenase activity comparable to that produced by insulin is readily

80 reconciled with the view that, insulin’s effect is mediated by Changes in the concentration of intramitochondrial metabolites rather than by a unique second messenger. Metabolites which are attractive candidates for this function include ATP, a substrate for pyruvate dehydrogena,se kinase, and ADP, a competitive inhibitor of pyruvate dehydrogenase kinase (11, 45). Severe depletion of cellular ATP levels clearly leads to pyruvate dehydrogenase activation (Table I). Possibly smaller changes in the mitochondrial ATP/ADP ratio occur in the presence of insulin, which go undetected when measuring total cellular nucleotides and yet are vital to the regulation of pyruvate dehydrogenase activity. Sccording to the calculations of Krebs and ‘Veech (48), the mitochondrial ATP/‘ADP .Pi ratio is sufficiently low so that small changes in this ratio could greatly affect the rate of pyruvate dehydrogenase phosphorylation despite the greater affinity of pyruvate dehydrogenase kinase for ATP (Km = 20 FM) than for ADP (K, = 100 PM) (45). Increasing the supply of fatty acids available for oxidation by providing them in the incubation medium might be expected to raise the mitochondrial ATP/ADP ratio, and this in turn could lead to the observed inactivation of pyruvate dehydrogenase. The effect of m-P-hydroxybutyrate can be similarly explained. Wieland and associates (39, 49, 50) have proposed a similar explanation for the effects of fatty acids and ketone bodies on the pyruvate dehydrogenase activity of heart, kidney, and liver. Some carboxylic acids, however, did not decrease pyruvate dehydrogenase activity, namely, pentanoate, acetate, propionate, and 4-pentenoate. In fact the latter three acids actually had an insulin-like effect in that they activated pyruvate dehydrogenase. Note that both acetate and propionate are converted to their CoA esters by enzymes found primarily in the cytosol (51) and that in adipose tissue the principle metabolic fate of cytoplasmic acetyl-Co$, and possibly propionyl-CoA as well, is conversion to long chain fatty acids (1, 52, 54). These processes of acid activation and lipogenesis would require ATP, and thus these acids might well diminish the mitochondrial ATP:ADP ratio rather than increase it and hence lead to pyruvate dehydrogenase activation. The activation of pyruvate dehydrogenase by 4-pentenoate might result from its ability to inhibit fatty acid oxidation (56). In addition, short chain carboxylic acids such as acetate and propionate are antilipolytic in adipose tissue (54,55) and thus might act in part by decreasing the supply of endogenous fatty acids available for mitochondrial oxidation. Hormonal regulation of pyruvate dehydrogenase activity appears to be of major significance in the control of lipogenesis in adipose tissue. In fact, the major function of pyruvate dehydrogenase in adipose tissue appears to be the provision of acetyl-CoA for synthesis of fatty acids via citrate (58). Bicarbonate ions thus are required both for the activation of pyruvate dehydrogenase by insulin and for the subsequent conversion of the acetyl-CoA produced to fatty acids. The intracellular mechanisms by which bicarbonate exerts its effect on pgruvate dehydrogenase activity remain obscure, but it is worth noting that the activation of pyruvate dehydrogenase by dinitrophenol, oligomycin, anoxia, and prolonged incubation with epinephrine are readily observed in Krebs-Ringer phosphate medium. Thus bicarbonate is apparently not involved in the activation process per se. Rather it seems to be required in order to achieve the proper concentration of mitochondrial metabolites essential for full pyruvate dehydrogenase activation. It is now quite apparent that pyruvate dehydrogenase represents a major rate-limiting step in the enzymatic sequence by

which pyruvate is converted to fatty acids (5, 8, 39). In fact we think it likely that pyruvate dehydrogenase activity determines the maximal rate of lipogenesis attainable in adipose tissue exposed to excess glucose and insulin. The fact that in this latter circumst,ancc the further addition of acetate increases glucose conversion to fatty acids (54,55,59) can now be simply accounted for by the activation of pyruvate dehydrogenase induced by acetate addition, although other more complex explanations can not be ruled out (60). This explanation for the effect of acetate on lipogencsis also accounts for the decrease in lact’ate release seen in the presence of acetate at a time when glucose uptake is increased (54). A puzzling feature of the action of niacin on adipose tissue from fasted-refed rats can now be given a partial explanation. It was found earlier that, whereas niacin accelerated fatty acid synthesis in this preparation in a manner similar to insulin, it was much less effective than insulin in inhibiting lactate release. This difference can now be attributed to the failure of niacin to block phosphorylase in tissue from refed rats. This observation raises intriguing questions with regard to the regulation of the hormone-sensitive lipase and glycogen phosphorylase. Acknowledgments-We have been greatly aided by the skilled technical assistance of Mrs. Maria Chung. Thanks are also due Dr. J. P. Flatt for valuable discussions and Dr. R. M. Denton for making available a manuscript prior to its publication. REFERENCES 1. WIXEGRAD, A. I., & RENOLD, A. E. (1958) b. Biol. Chem. 233. 267-272 2. LEVINE, R., GOLDSTEIN, M., KLEIN, S., & HUDDLESTUN, B. (1949) J. Biol. Chem. 1’79, 985-9% 3. FOSTER, D. W., & BLOOM, B. (1963) J. Bio2. Chem. 238, 888892 4. JUNGAS, R. L. (1968) Biochemistry 7, 3708-3717 5. JUNQAS, R. L. (1970) Endocrinology 86, 1368-1375 6. JUNGAS, R. L. (1971) Metabolism 20,43-53 7. DENTON, R. M., COORE, H. G., MARTIN, B. R., & RANDLE, P. J. (1971) Nature New Biol. 231,115-116 8. COORE, H. G., DENTON, R. M., MARTIN, B. R., & RANDLE, P. J. (1971) Biochem. J. 125, 115-117 G., SCHIRMANN, A., & WIELAND, 0. 9. WEISS, ‘L., ~FFLER, (1971) Fed. Eur. Biochem. Sot. Lelt. 16, 229-231 10. LINN, T. C., PETTIT, F. H., & REED, L. J. (1969) Proc. Nut. Acad. Sci. U. S. A. 62,234241 11. LINN, T. C., PETTIT, F. II., HUCHO, F., & REED, L. J. (1969) Proc. Nut. Acad. Sci. U. S. A. 64.227-234 12. TAYLOR, S. I., & JUNGAS, R. L. (1972) Fed. Proc. 31,244 13. JUNG&R. L.; & T2k~1,0~, S. I. (1972) inInsulin A&n (FRITZ,

I. B., ed) pp. 369-413, Academic Press, New York 14. KREBS, H. A., & HENSI~XZIT, K. (1932) Z. Physiol. Chem. 210, 33-66 15. UMBREIT, W. W., BUXRIS, IL. II., & STAUBFI~;R, S. F. (1957) Manometric Techniques, 3rd Ed, pp. 149-150, Burgess

Publishing

Co., Minneapolis

16. JUNGAS. It. L. (1966) Proc. Nat. Acad. Sci. U. S. A. 66.757-763 17. DOLE, $. P., &‘ ME&EETZ, II. (1960) J. Biol. Chem. 236,25951 2599 18. HALPERIN, M. L., & DENTON, R. M. (1969) Biochem. J. 113, 207-214 19. FROESCH, E. R. (1965) in Adipose Tissue (RENOLD, A. E., $ CAHILL, G. F., JR., eds) pp. 281-293, Waverly Press, Balti-

more 20. JUNGAS, 21. MAHLER,

R. L., & BALL, E. G. (1963) Biochemistry 62, 383-388 R., STAFFORD, W. S., TARRANT, M. E., & ASHMORE, J. E. (1964) Diabetes 13,297-302 22. BUTCHER R. W., SNEYD, J. G. T., PARK, C. R., & SUTHERLAND, E. W., JR. (1966) J. Biol. Chem. 241, 1651-1653 23. Kuo, J. F., & DE RENZO, E. C. (1969) J. Biol. Chem. 244, 2252-2260 24. CARLSON, L. A. (1963) Acta Med. Stand. 173.719-722

81 R. W., BAIRD, C. E., & SUTHERLAND, E. W. (1968) Chem. 243, 1705-1712 26. GITRRITSICN, G. C., & DIJLIN, W. X. (1905) J. Pharmacol. Exp. Ther. 160, 491-498 27. BUTCHICR, 12. W. (1970) in Adipose Tissue: Regulation and JZetabolic Function (HXPP, D., & JEANRENAUD, B., eds), pp. 5-10, Academic Press, New Tork

25.

BUT~HI~I~,

J. Biol.

28. 29.

STEIXI~ERG, D., VAUGH.\N, &I., XIXTIX,, P. J., STRAND, O., & BERGSTROM, S. (1964) J. Clin. Znvest. 43, 1553-1540 BCTCHCR, R. W., & BAII~D, C. E. (1908) J. Biol. Chem. 243,

1713-1717 30. CLTXAND, W. W. (19G7) A&an. Enzymol. 29, 1-32 31. SIESS, E., WITTMAN, J., & WILLAND, 0. (1971) Hoppe-Seyler’s 2. Physiol. Chem. 362,44i-452 32. SCHNAITMAN, C., & GREIXAVILT, J. W. (1968) J. Cell Biol. 38, 158-175 33. Looms, W. F., & LIP~~ANX, F. (194s) J. Biol. Chem. 173,807808 34. LARDY, H. A., JOHNSON, P., & RICRIURRAY, W. C. (1958) Arch. Biochem. Biovhus. 78. 587-597 35. EDTARDS, S. W., & ‘B-ILL, E. G. (1954) J. Biol. Chem. 209, GlQ-633 36. PRESSMAN, B. C., & LARDY, H. A. (1956) Biochim. Biophys. Acta 21, 458-466 37. HEPP, D., CHALLONER, D. II., & WILLIAMS, R. H. (1968) J. Biol. Chem. 243, 2321-2327 38. BIHLER, I., & JEA&RENAUD, B. (1970) Biochim. Biophys. Acta 202, 496-506 39. WIELAND, 0. H., PATZELT, C., & L~FFLER, G. (1972) Eur. J. Biochem. 26, 426-433 40. KNEER, P., & BILL, E. CT. (lQG8) J. Biol. Chem. 243, 28632870 41. C~ATRECASAS, P. (1969) Proc. NT&. Acad. Sci. U. S. A. 63, 450-457 42. CUATRIXASAS, P. (1970) Proc. Sat. Acad. Sci. U. S. A. 68, 1264-1268 43. Ho, R. J., JEANRENAUD, B., POSTERNAIC, T. H., & RENOLD, A. E. (1967) Biochim. Biophys. Acia 144,74-82

44. MARTIN, B. R., DENTON, R. M., PASIC, H. T., & RANDLE, P. J. (1972) Biochem. J. 129, 763-773 45. HUCHO, F., R.~ND~LL, D. D.? ROCHE, T. E., BURGI~TT, M. W., PELLISI-, J. W., & REED, L. J. (1972) Arch. Biochem. Biophys. ltil, 328-340 ’ ’ 46. WIELAND, O., & Smss, E. (1970) Proc. Nat. Acad. Sci. U. S. A. 66,947-954 47. ScHrxnuiL, R. J., & GOODMAN, 1%. n!I. (1972) Biochim. Biophys. Acfa 260, 153-158 48 K&q H. A., & VEIXH, R. L. (1969) in The Energy Level and Metabolic Control in Mitochondria (PAPA, S., TASER, J. M., &UAGLIARIELLO, E., & SLATER, E. C., eds) pp. 329-382, Adriatica Editrice, Bari 49. WIELAND, O., FUNCKI”, H. v., & L~FFLER, G. (1971) Fed. Eur. Biochem. Sot. Lett. 15, 295-298 50. WIELAXD, O., Smss, E., SCHULTZE-WETHMAR, F. H., FUNCKE, H. G. v., & WIXTON. Bionhus. , B. (1971) Arch. Biochem. 1 Y 143. 593-601 51. MARTIN, B. R., & DENTON, R. M. (1970) Biochem. J. 117,861877 52. FLATT, J. P., & BILL, E. G. (1964) J. Biol. Chem. 239, 675485 53. LANDAU, B. R., & KATZ, J. (1964) J. Biol. Chem. 239, 697-704 54. FLATT, J. P., & BALL, E. G. (1964) J. Biol. Chem. 241, 28622869 55. DELBOCA, J. (196s) Effets de L’Acetate Sur Le Metabolisme In Vitro clu Tissu Adipeux Epididymaire, Imprimerie Nestle, Echandens 56. BRESSLER, R., CORREDOR, C., & BRENDIXL, K. (19GQ) Pharmacol. Rev. 21, 105-130 57. BRINGOLF, M., ZARAGOSA, N., RIVIER, D., & FELBERS, 7. P. (1972) Eur. J. B&hem. 26, 360-367 58. KORNACKER, M. S., & BALL, E. G. (1965) Proc. Nat. Acad. Sci. U. S. A. 64, 899Q904 59. DELBOCA, J., & FLATT, J. P. (1969) Eur. J. Biochem. 11, 127134 f-50.FLATT, J. P. (1970) J. Lipid Res. 11, 131-143