the physiological role of pyruvate carboxylation in hamster brown adipose tissue

Eur. J. Biochem. 94, 419-426 (1979)

The Physiological Role of Pyruvate Carboxylation in Hamster Brown Adipose Tissue Barbara CANNON and Jan NEDERGAARD

The Wenner-Gren Institute, University of Stockholm

(Received July 5/0ctober 27, 1978)

1. Pyruvate carboxylase is present in brown adipose tissue mitochondria. 2. In isolated mitochondria, pyruvate, bicarbonate and ATP, the substrates for pyruvate

carboxylase, are able to replace added malate in supplying a condensing partner for acetyl-CoA formed from P-oxidation of fatty acids.

3. In brown adipocytes, pyruvate and C02 increase the rate of norepinephrine-stimulated respiration synergistically.

4. The norepinephrine-stimulated respiration in brown adipocytes is diminished when pyruvate transport into the mitochondria is inhibited.

5. Pyruvate carboxylation increases the intramitochondrial level of citric acid cycle intermediates, as shown by titrations of malonate inhibition of respiration.

6. Pyruvate carboxylation can continuously supply the mitochondria with citric acid cycle intermediates, as evidenced by its ability to maintain respiration when oxoglutarate conversion to glutamate is stimulated.

7. Pyruvate carboxylation is necessary for maximal oxygen consumption even when drainage of the citric acid cycle for amino acid synthesis is eliminated.

8. Pyruvate carboxylation explains observed effects of COz on respiration in brown adipocytes, and may also explain the increased glucose uptake by brown adipose tissue during thermogenesis in vivo.

Isolated brown adipocytes from hamster respond to the neurotransmitter norepinephrine with a 20- 40-fold increase in respiratory rate [l]. This increased respiration in vitvo is a manifestation of the heat pro- duction in brown adipose tissue [2,3] which occurs when a hamster is exposed to cold. In earlier experiments, the respiratory increase was of a transitory nature [l]. However, it was recently observed in our laboratory that bubbling of the Krebs/Ringer/phosphate buffer with 5 % C02 in air resulted in a norepinephrine-stimulated respiratory rate which re- mained constant with time. Not only was this respiratory rate constant, but it was also at least double as high as that seen in the absence of C02 [4]. Similarly, the respiratory rate in Krebs/Ringer/bicarbonate buf-

Abbreviation. Tes, N-tris(hydroxymethyl)methyl-2-aminoeth- ane sulphonic acid.

Enzyme. Pyruvate carboxylase or pyruvate : carbon-dioxide ligase (ADP-forming) (EC 6.4.1.1).

Note. In this paper, utilisation of the terms carbon dioxide and bicarbonate denotes the species added, and does not imply any specification of the active species.

fer (+ albumin) is stable with time, and is higher than in the phosphate + C02 buffer [5].

Carbon dioxide and bicarbonate can have a number of effects on cellular metabolism. Firstly, bubbling with C02 decreases the pH of the phosphate buffer to 6.95 which may alter the intracellular pH [4]; but since the pH of the bicarbonate buffer is 7.4 it is unlikely that this acidification could be responsible for the beneficial effects of C02. Secondly, in the absence of C02/HCO?, the cellular membrane potential is lower than in the presence [6]. Thirdly, bicarbonate may become fixed, either in anabolic processes in the cytoplasm, or in anaplerotic reactions for the mitochondrial citric acid cycle. In biochemical thermogenesis a role for anaplerotic reactions can be readily visualised.

The substrate for thermogenesis is fatty acids, released in large amounts upon norepinephrine stim- ulation. These are partially oxidised in mitochondrial p-oxidation to acetyl-CoA. For complete oxidation, acetyl-CoA must condense with oxaloacetate to enter the citric acid cycle. If this step is not to become

420 Pyruvate Carboxylation in Brown Adipose Tissue

rate-limiting, oxaloacetate must be presented to citrate synthase as rapidly as acetyl-CoA is produced.

The transition from the resting to the maximally stimulated respiratory rate occurs in a period of one to two minutes. The 40-fold increase in metabolic rate which occurs in this short time interval demands a large increase in the capacity of the citric acid cycle. This demand may be partially fulfilled by an increased rate of turnover of the citric acid cycle, and, more directly, by increasing the amount of oxaloacetate in the mitochondria.

The enzyme pyruvate carboxylase could supply more oxaloacetate by catalysing its formation from pyruvate and bicarbonate.

Pyruvate cdrboxyhse [7] was first demonstrated in animal tissues by Utter and Keech [8- lo]. Shortly after this, Wise and Ball demonstrated its occurrence in white adipose tissue [ll]. In animal tissues, it has been suggested to play an important role in gluco- neogenesis, in lipogenesis and in the synthesis of glutamate (see eg. [12-151). The enzyme is localised to the mitochondria [16,17] and in isolated mitochondria the reaction products are found as malate and citrate [14,18]. A review concerning the conversion of pyruvate to oxaloacetate in various species has appeared recently [19].

The aim of the present study has been to examine if pyruvate carboxylase is present in brown adipose tissue mitochondria; if so, if it can support fatty acid oxidation by supplying oxaloacetate, and finally to investigate if this is a likely explanation for the observed effect of COz/HCO: on norepinephrine-stim- dated respiration in brown adipocytes. Some of these results have been presented elsewhere [20 - 221.

MATERIALS AND METHODS

Brown Adipose Tissue

Brown adipose tissue mitochondria were isolated, as previously described [23], from the pooled brown adipose tissue of golden hamsters (Mesocricetus aura- tus), which had been cold-adapted at 5 "C for at least three weeks with food and water ad libitum. The mitochondria were suspended finally in 100mM KCI containing 20 mM K-Tes, pH 7.2.

Brown adipocytes were isolated from the pooled brown adipose tissue of golden hamsters which were maintained at 22 "C [5,24]. The number of cells per ml was determined by counting a dilute suspension in a Burker chamber [24]. The fraction of broken cells was estimated by staining with 0.1 % Alcian blue. This fraction was routinely less than 10 %.

Pyruvate Carboxyluse Activity

Pyruvate carboxylase activity was measured in isolated mitochondria as incorporation of radioactive

bicarbonate into acid-stable products, approximately as described by Pate1 and Hanson for white adipose tissue mitochondria [25]. The basic reaction medium consisted of 100mM KC1, 20mM K-Tes, 10mM potassium phosphate, 20 mM MgC12 and 1 mM EDTA at pH 7.6. The medium also contained 40 mM KHCOJ, 0.4 pM carbonylcyanidep-trifluoromethoxyphenylhydrazone, 10 mM pyruvate, 10 mM ATP, 0.6 mM arsenite, 5 mM malonate and 50 pM fluorocitrate. Radioactive H[14C]03 was added to a specific activity of 5 - 10000 counts x min-' x pmol-'. Mito- chondria were added to a final concentration of 3 mg protein/ml, and the total incubation volume was 1 ml. The mixture was preincubated for 5 min at 37 "C. After this, 5 mM acetyl-L-carnitine was added, and the incubation continued for a further 30 min. The reaction was terminated by transferring the incubation tubes to ice and adding perchloric acid to a final concentration of 6 %. To remove unfixed radioactive bicarbonate, the mixture was bubbled for 4 min with 100% C02. The tubes were centrifuged at 15OOxg for 10 min and an aliquot of the supernatant was taken to determine the acid-stable radioactivity. When the influence of pH was studied, i t was found necessary to use 1 mM phosphate and 20 mM bicarbonate in order to obtain the pH range required.

The Product of Bicarbonate Fixation

The product of bicarbonate fixation was investigated on samples of the above supernatant which had been neutralised by K2C03. The samples were applied onto precoated cellulose thin-layer chromatog- raphy plates (Merck). The developing system was di- ethylether/formic acid/water (1 50/110/40). Radioac- tive malate and citrate were used as standards. After development, 5-mm bands were scraped off the plates for determination of radioactivity.

Radioactivity

Radioactivity was determined in a scintillation fluid consisting of 0.5 % (w/v) 2,5-diphenyloxazole dissolved in toluene/Triton X-100 (2/3). The samples were counted in an Intertechnique SL36 liquid scintillation spectrometer.

Oxygen Consumption

Oxygen consumption was measured using a Yellow Springs Instrument 4004 Clark Oxygen Probe; both concentration and rate of utilisation of oxygen were monitored. The chamber volume was 1 ml and the temperature 23 "C for mitochondria1 studies and 37 "C for experiments with adipocytes. The final concentration of mitochondria was 0.7 mg protein per ml, and of adipocytes 100000 cells per ml. Other details as in legends to figures.

B. Cannon and J . Nedergaard 42 1

Protein

Protein was determined by the biuret method [26], with bovine serum albumin as standard.

Chemicals

Palmitoyl-L-carnitine and acetyl-L-carnitine were gifts from Otsuka Pharmaceutical Co. (Osaka, Japan). Compound UK5099 [a-cyano-P-(l-phenylindol-3-yl)- acrylate] was a gift from Pfizer (Sandwich, England). Carbonylcyanide p-trifluoromethoxyphenylhydrazone was purchased from Pierce Eurochemie (Rotterdam, Holland), crude collagenase from Sigma (St Louis, Mo., U.S.A.) and fatty acid free bovine serum albumin, fraction V, from Miles (Kankakee, Ill., U.S.A.). Radioactive chemicals were from the Radiochemical Centre (Amersham, England). All other compounds were of the highest purity commercially available.

RESULTS AND DISCUSSION

Conditions Required for Hf ' 'C/0; Fixation by Brown Adipose Tissue Mitochondria

If pyruvate carboxylase is present in brown adipose tissue mitochondria, bicarbonate should be incorpo- rated into acid-stable compounds, and this incorporation should be dependent upon the presence of both pyruvate and ATP.

ATP is not synthesised by oxidative phosphorylation in these isolated mitochondria, as brown adipose tissue mitochondria are uncoupled when convention- ally prepared [27]. They were used in this state for the present investigation.

In order to prevent citric acid cycle metabolism and thus loss of fixed bicarbonate as C02 during the incubation period, malonate and fluorocitrate were added to the reaction medium. Arsenite, at a concentration which entirely inhibited pyruvate oxidation, was also present. This facilitated an investigation as to whether pyruvate carboxylase in brown adipose tissue mitochondria required acetyl-CoA as a positive effector, as has been previously reported for other tissues [8,9,28 - 301.

Brown adipose tissue mitochondria were found to be capable of bicarbonate fixation in the incubation medium used with an activity at 37 "C of 35-40 nmol per min per mg mitochondrial protein. The dependence of this fixation upon mitochondrial protein concentration was studied and found to be linear at least up to 3 mg per ml. The reaction rate was constant with time up to 30 min. Routinely 3 mg mitochondrial protein per ml and an incubation time of 30 min were used.

The optimal concentrations of substrates and co- factors for H[14C]O; fixation were determined and they are shown in Table 1.

Table 1, The concentrations of substrates and cofuctors required ,fbr H["C]O; fixation in hrown adipose tissue mitochondriu For conditions, see Materials and Methods

Substrate or cofactor Concentration Optimal for half-maximal concentration effect

mM

Pyruvate 0.1 3

MgZ + 2.5 10-20 Acetyl-carnitine 0.1 1

Bicarbonate 14 25 ATP 2.5 10-20

The optimal concentration for bicarbonate is in agreement with the concentration usually found in blood (26-28 mM), and may thus be in the physiological range. The dependence of H['4C]O; fixation upon acetyl-carnitine must be interpreted as a require- ment of pyruvate carboxylase for acetyl-CoA, this being readily formed from acetyl-carnitine in these mitochondria 1311. In other animal systems acetyl- CoA also acts as a positive effector [8,9,28 - 301.

No dependence of H['4C]0, fixation upon the presence of phosphate could be found, when phosphate was titrated up to 30mM. Such a dependence has, however, been recorded for white adipose tissue [25].

The pH dependence of H['4C]O; fixation in brown adipose tissue mitochondria was also investigated. Maximum activity was achieved above pH 7.4 and was constant to pH 8, the highest pH tested. It is feasible that when the pH is decreased below 7.4, the concentration of bicarbonate could become rate- limiting as C02 is lost.

Mehlman el al. reported that ADP is inhibitory to pyruvate carboxylase from kidney [12]. H[14C]0, fixation in brown adipose tissue mitochondria was also inhibited by ADP, from 1 mM concentration, even though the reaction was carried out in the presence of 10 mM ATP. At this ATP concentration, 20 mM ADP was required for half-maximal inhibition. It is thus possible to conclude that brown adipose tissue mitochondria do contain pyruvate carboxylase. The activity found was always in the range 35-40 nmol per min per mg mitochondrial protein at 37 "C. This activity is in the same range as that found in white adipose tissue [25] and in kidney [12].

The Product of Bicarbonate Fixation

When the citric acid cycle is active, the fixed bicarbonate will eventually be released again as COZ. However, in the presence of malonate and fluorocitrate, the H[14C]O; fixed by pyruvate carboxylase is expected to be found as malate and citrate [14,18]. This was studied in an extract of brown adipose tissue


Pm Cn

125nmo‘ ‘ ‘ L C O ; --. ,ATP

H + Malate. lmin

Fig. 1. Effect of p ~ ~ u v a t e carho.yjlutioti upoii the extent oflfatty acid oxidation in brown adipose tissue mitochondria. 0.7 mg mitochondria1 protein were added to 3 ml of a medium consisting of 100 mM KCI, 20 mM K-Tes, 2 mM MgC12, 4 m M KH2P04 and 1 mM EDTA. Additionally carbonyl cyanide p-trifluoromethoxy hy- drazone (0.4 pM) and 0.6 mM As203 were present. 12 nmol palmitoyl-carnitine (PmCn), 3 mM malate, 3 mM pyruvate, 1.5 mM ATP and 6 mM KHCO3 were added where indicated

mitochondria after fixation, using thin-layer chroma- tography. No radioactivity from the sample migrated together with malate. All the radioactivity migrated in parallel with the citrate standard. The product of H [‘“CIO, fixation was therefore tentatively identified as citrate. The mitochondria were completely uncoupled during the incubation and had thus a low concentration of NADH. It is consequently not sur- prising that no malate was identified in the reaction products, since NADH is required for the formation of malate from oxaloacetate. Even when the incubation was carried out in the presence of 3 mM malate to serve as a trap for any radioactive malate formed, no radioactivity was found to migrate with standard malate.

Pyruvate Carboxylation and Fatty Acid Oxidation in Brown Adipose Tissue Mitochondria

In order to postulate a role for pyruvate carboxylase in fatty acid oxidation in brown adipose tissue, it is necessary to demonstrate that pyruvate carboxylation can provide oxaloacetate, required for conden- sation with acetyl-CoA formed in P-oxidation, at an adequate rate. When fatty acid oxidation is studied in isolated mitochondria, malate is generally added to supply oxaloacetate. Under conditions where pyruvate carboxylase is active, malate addition should therefore be superfluous.

Brown adipose tissue mitochondria from hamster are able to oxidise fatty acid to free acetate in the absence of added malate or carnitine [32]. This partial oxidation of added fatty acid, here a limited amount of palmitoyl-carnitine, consumes a certain amount of oxygen (Fig.1). When malate is present

Table 2. Synergistic effect of pyruvate and carbon dioxide on mmi- ma! rate of respiration in brown adipocytes Cells, prepared, stored and incubated without added glucose and fructose, were incubated at a concentration of lUO000 cells per rnl in Krebs/Ringer/phosphate buffer [ S ] . Pyruvate (10 mM) was present in the incubation medium where indicated. Also where indicated the medium was bubbled with 5 COZ in air Basal respiratory rate was measured over 5 niin prior to addition of 1 pM norepinephrine. Values are mean S.E.M. (n = 6). The data marked with an asterisk (*) are significantly different from each other (P < 0.05, paired observation, t-test, n = 6 )

Respiratory rate Basal respiratory Increase due affected by rate to norepinephrine

No addition Pyruvate

COZ Increase due to pyruvate

Increase due to C02 Expected increase due to pyruvate + COz

Observed increase due to pyruvate + C02

COZ + pyruvate

nmol O x min-’ x 106 cells-’

4 8 k 6 190 k 14 221 k 19 13k 8

2 6 k 8 31 k 1 1 4 9 1 5 502 * 22

5 1 5 311 * 25

~ _ _ _

26 k 12 6 0 k 4 585 21

1 3 k 16

341 2 25*

394 k 30*

in the incubation medium, then the amount of oxygen consumed during oxidation of the same amount of palmitoyl-carnitine is increased due to entry into the citric acid cycle. (That the added palmitoyl-carnitine is exhausted when the oxygen consumption returns to a low rate has been previously demonstrated in this laboratory using [‘4C]palmitoyl-carnitine [32], also that the extra oxygen consumption is due to entry into the citric acid cycle [31,32].) When malate was replaced by pyruvate, bicarbonate and ATP, these three together also facilitated the increased oxygen consumption. All three components were necessary to observe this effect. It thus appears that under conditions where pyruvate carboxylase is active, oxaloacetate is delivered to the citric acid cycle at a rate sufficient to maintain fatty acid oxidation at the same rate as when malate is used to supply oxaloacetate.

Since arsenite was present in these experiments at a concentration which completely inhibited pyruvate dehydrogenase, oxoglutarate dehydrogenase was also inhibited, which prevented complete oxidation of the fatty acid and thus limited the ox) gen consumption.

Synergistic Effect of Pyruvate and Carbon Dioxide in Brown Adipocytes

As demonstrated, pyruvate carboxylase is able to provide a condensing partner for acetyl-CoA in isolated mitochondria. As an indication of the activity of pyruvate carboxylase in isolated brown adipocytes,

B. Cannon and J . Nedergaard 423

a synergistic effect of pyruvate and CO, on norepinephrine-induced respiration should be found. Table 2 presents the results of experiments using brown adipocytes (stored and incubated without added glucose and fructose) where the norepinephrine-stimulated respiratory rate was maximal when both pyruvate and carbon dioxide were present in the incubation medium. The increase found with pyruvate plus COZ is significantly greater than that expected from the sum of the two separate additions ( P < 0.05, paired observation, t-test, n = 6). This is an indication that pyruvate carboxylase can be active in the isolated cells. Although the cells are suggested to be loosely coupled during norepinephrine-stimulated respiration, ATP, needed for pyruvate carboxylation, is presumably produced both by some respiratory- chain-linked phosphorylation and by obligatory substrate-level phosphorylation. The influence of added pyruvate is difficult to observe if glucose and fructose are present in the medium, since they will give rise to pyruvate via glycolysis.

Effect on Oxygen Consumption in Brown Adipocytes of Inhibition of Pyruvate Permeation into the Mitochondria

Halestrap and Denton have reported a series of cinnamic acid derivatives which, at low concentrations, specifically inhibit the mitochondrial pyruvate carrier [33 - 351. These compounds were also shown to be effective in intact tissue preparations.

In an attempt to obtain evidence for the action of pyruvate carboxylase in isolated brown adipocytes in the presence of carbon dioxide, the influence of one of these inhibitors, compound UK5099, on norepinephrine-stimulated respiration has been studied. We have confirmed that compound UK5099 can inhibit pyruvate transport also in brown adipose tissue mitochondria. Since pyruvate carboxylase is a mitochondrial matrix enzyme, and pyruvate is produced cytoplasmically by glycolysis, then inhibition of the mitochondrial pyruvate carrier would be expected to inhibit pyruvate carboxylation due to pyruvate depletion. Under conditions where pyruvate carboxylase is not active, inhibitors of pyruvate transport should be without effect.

In Fig.2A is shown the action of compound UK5099 upon the rate of norepinephrine-stimulated respiration in brown adipocytes in buffer bubbled with COZ (i.e. when pyruvate carboxylase should be active). The respiratory rate was halved. A high concentration of pyruvate added to the inhibited system was able to stimulate respiration up to approximately the original rate. (Pyruvate addition to the uninhibited system was without effect.) In Fig. 2B is depicted the rate of norepinephrine-stimulated respiration in the absence of COZ (i.e. when pyruvate

OOnmoi

n n

py ruva t e H lrnin

NE

B

I H t 1 rnin

NE

Fig. 2. Influence of compound UKSOYY 011 riol.c,/~iiieplirine-srimulared respiration in brown adipocytes. The incubation medium in the upper traces was Krebs/Ringer/phosphate buffer bubbled with S "/, COZ in air. In the lower traces the medium was Krebs/Ringer/phosphate buffer. NE: 1 pM norepinephrine; UK: 10pM compound UKS099; pyruvate: 10 mM pyruvate. Other conditions as described in Mate- rials and Methods

carboxylase should be inactive), and the lack of effect of compound UK5099.

Cyanocinnamates, at high concentrations, may be inhibitory to translocation of carnitine and acyl- carnitine across the mitochondrial membrane [36] but it is unlikely that compound UK5099 could cause such an inhibition at the low concentration used here; additionally, in the absence of COz, no inhibition occurred (Fig.2B). The effect of pyruvate in Fig.2A may be explained by the suggestion that pyruvate, at high concentrations, is able to diffuse across the inner membrane as the undissociated acid, thus circumvent- ing of inhibition of its carrier [37], and allowing ge- neration of oxaloacetate.

These experiments support the contention that transport of pyruvate into the mitochondrial matrix is occurring during norepinephrine-stimulated respiration, at least in the presence of carbon dioxide, and that this is necessary for maintenance of the increased respiratory rate.

Indications of a Higher Level of Intramitochondrial Citric Acid Cycle Intermediates due to Pyruvate Carboxylation

If pyruvate carboxylation is able to increase concentrations of citric acid cycle intermediates in the cells, experiments may be performed which provide indirect evidence for the presence of these higher levels when carbon dioxide and pyruvate are present. Fig.3A shows an example of how these experiments were performed. The influence of varying concentrations of malonate upon norepinephrine-stimulated


NE rnalonate

i 4

A

H lmin

100

- - .- 2 E ,g c 0 .- c m E

5 50

.- a w

n

- .- i 1 w

a, c

r a a, c a

.- L

.- 2 z

0 0 10 20

Malonate (mM)

- -0 O\

\

C

L. Fig. 3. Influence of mulonate on norepineplirine-stimuluie(~ rrspirution in brown adipocytes. (A) An example of' malooate inhibition of norepinephrine-stimulated respiration. NE : 1 pM norepinephrine. Here, the medium was Krebs/Ringer/phosphate buffer bubbled with 5 % C02 in air; malonate: 5 mM. (a) Maximum of norepinephrine-stimulated respiratory rate. (b) Respiratory rate 1 niin after malonate addition. (c) Respiratory rate 5 min after malonate addition. (B) Comparison of malonate inhibition in KrebslRingeriphosphate buffer with and without COz. Results of a series of experiments performed as in A. Open symbols: bja, 1 min; closed symbols: cia, 5 miii. (0,m) With C 0 2 ; (o,.) no C 0 2 added. (C) Comparison of malonate inhibition in Krebs/Ringer/bicarbonate buffer with and without 10 mM pyruvate. 4 % fatty-acid-free bovine serum albumin was present. Experiments performed as in A. Open symbols: bjd, 1 min; closed symbols: cja, 5 min. (0, .) With pyruvate; (O,.) no pyruvate added

1100 nmol O/ min t NE

Fig. 4. Eflect o j ammonia on norepinephrine-stimulated respiration in brown adipocytes. The incubation medium in the upper traces was Krebs/Ringer/phosphate buffer bubbled with 5 "/, COZ in air. In the lower traces the medium was Krebs/Ringer/phosphate buffer. NE: 1 pM norepinephrine. NH3: 10 mM NH,C1 was added at the arrow

respiration at two time intervals after inhibitor addition has been studied. In Fig.3B a comparison is made between the inhibitory effect of malonate in the presence and in the absence of C02 in Krebs/ Ringer/phosphate buffer. It can be seen that in the presence of C02, higher malonate concentrations are required to obtain the same degree of inhibition than in the absence of C02.

Since malonate is a competitive inhibitor of succinate dehydrogenase, this implies that the concentration of succinate is higher when C02 is present, i.e. when pyruvate carboxylase is expected to be active.

Fig. 3C depicts the results of similar experiments to those reported in Fig. 3 B, but here the comparison is made between the presence and the absence of pyruvate in Krebs/Ringer/bicarbonate buffer. In the presence of pyruvate, when pyruvate carboxylase is probably stimulated, higher malonate concentrations are required to obtain the same degree of inhibition than in the absence of pyruvate.

These experiments with malonate inhibition are indicative of an increased level of intramitochondrial intermediates as an effect of pyruvate carboxylation.

Continuous Supply of Citric Acid Cycle Intermediates by Pyruvate Carboxylase in Brown Adipocytes

In Fig. 4 the lower trace depicts norepinephrine- stimulated respiration in Krebs/Ringer/phosphate buffer. At the arrow, 10 mM ammonium chloride was added. N o measurable pH change occurred. The respiration immediately ceased. In the upper trace, where the buffer had been bubbled with 5 % C02, the effect of ammonium chloride was much less pro- nounced. The rate of respiration did decrease, but only slowly. The presence of ammonium will stimulate

B. Cannon and J. Nedergaard 425

glutamate synthesis which, in the absence of COz, can be expected to deplete the pool of citric acid cycle intermediates by converting 2-oxoglutarate to glutamate. In the presence of COZ, pyruvate carboxylase will continuously compensate for the leak of intermediates by synthesising oxaloacetate. That respiration decreases at all is probably due to the limited activity of pyruvate carboxylase.

When the same experiment was performed in the presence of both 10 mM pyruvate and COZ, ammonium chloride produced no depression of norepinephrine-stimulated respiration (not shown). When a saturating concentration of pyruvate is ensured, pyruvate carboxylase can apparently maintain a rate of oxaloacetate synthesis which is at least equal to the rate of removal of oxoglutarate by glutamate dehydrogenase. When pyruvate was added to the incubation performed in the absence of COZ, it was not able to prevent the inhibition by ammonium chloride. This thus eliminates the possibility that the action of pyruvate in relieving the ammonium chloride inhibition could be via a transamination reaction which would regenerate oxoglutarate : pyruvate + glutamate S 2-oxoglutarate -t aspartate.

When the same experiment as in Fig.4 was performed using only 1 mM ammonium chloride the respiration in the absence of COZ ceased as in the figure; but, in the presence of COZ, no inhibition developed. Apparently the rate of removal of oxoglutarate was limited in this case, as would be expected in view of the high K, (3 mM) of glutamate dehydrogenase for ammonium [38].

The Nitrogen Balance of Isolated Brown Adiyocytes

A plausible explanation for the decline of norepinephrine-stimulated respiration in adipocytes incubated without C 0 2 may be a leak of citric acid cycle intermediates. From the above, the possibility arose that the brown adipocytes might be in negative nitrogen balance after isolation and storage, and that amino acid synthesis might be the cause of such a leak of citric acid cycle intermediates. To investigate this, two preparations of brown adipocytes were made from one hamster: one was made in the routine way, while the other was prepared with a mixture of 20 amino acids in the medium. The cells were then tested in buffer with and without COZ. The cells prepared in the amino acid mixture were incubated in the presence of the amino acids. The results of such an experiment are shown in Fig.5. In the absence of COZ, amino acids prevented the decline in stimulated respiratory rate, but were not, however, capable of increasing the respiratory rate to that seen with COZ. When COZ was present, amino acids had no obvious effect on respiration.

+COP +amino ac ids r I

+Amino acids

50nmol Oimin lmin

Fig. 5. Ej’ect qf unzino acids on norepinephrine-stimulute~ rrspira- tion in brown udipocytes. The incubation medium was Krebs/ Ringer/phosphate buffer. + amino acids: a balanced mixture of the 20 amino acids (2 mM total concentration) was added to the storage and incubation media. + COZ : buffer bubbled with 5 COZ in air. N E : 1 pM norepincphrine

It is thus feasible that amino acids are lost during cell isolation and storage, and that this loss manifests itself after norepinephrine addition when citric acid cycle intermediates are apparently used to synthesise amino acids. Incubation with amino acids compen- sates for any such losses and consequently allows a constant rate of norepinephrine-stimulated respiration. Maximum rates of oxygen consumption demand, however, that the level of citric acid cycle intermediates is increased above the level in the resting cell. This can be achieved in the presence of COZ/ HCO; by pyruvate carboxylase.

CONCLUSIONS

On the basis of the results presented here it is possible to conclude that the beneficial effect of C 0 2 on the rate of oxygen consumption (= thermogenesis) in brown adipocytes is due to the activity of pyruvate carboxylase.

The maximum oxygen consumption rates during thermogenesis demand that the level of citric acid cycle intermediates is increased above the level in the resting cell. Pyruvate carboxylase accomplishes this, and heat evolution is doubled when pyruvate carboxylation takes place.

These results may also explain earlier observations from experiments in vivo, where glucose uptake in brown adipose tissue is stimulated when thermogenesis takes place [39,40]. This glucose, in the form of pyruvate, may be necessary for pyruvate carboxylation.

The situation observed here in brown adipose tissue resembles that observed in insect flight muscle when an insect takes to flight. The metabolic rate, and hence the activity of the citric acid cycle, also there increases up to 50 times [41]. In order to accommodate this

426 B. Cannon

large increase in rate of the citric acid cycle, Crabtree et al. [41] proposed that oxaloacetate must be provided from reactions outside the cycle. They suggested that pyruvate should be the source of the extra oxaloacetate and that the conversion of pyruvate to oxaloacetate was catalysed by pyruvate carboxylase. This role of pyruvate carboxylase has been termed ‘internal anaplerotic’, since the oxaloacetate is not required for synthetic reactions [42].

Brown adipose tissue seems to provide another example of this role of pyruvate carboxylase.

Competent technical assistance was provided by Agneta Berg- strom and Barhro Kyte. The work was supported by the Swedish Natural Science Research Council (grant B2171-028).

REFERENCES

1. Prusiner, S. B., Cannon, B. & Lindberg, 0. (1968) Eur. J . Bio-

2. Smith, R . E. & Horwitz, B. A. (1969) Physiol. Rev. 49, 330-

3. tindberg, 0. (1970) Brown Adipose Tissue, Elsevier, New York. 4. Pettersson, B. ( I 977) Eur. J . Biochem. 72, 235 - 240.

chem. 6, 15-22.

425.

5.

6.

7.

8. 9.

10.

11.

12.

13. 14. 15.

16.

17.

Nedergaard, J. & Lindberg, 0. (1979) Eur. J . Biochem. in the press.

Williams, J . A. & Matthews, E. K. (1974) Am. J . Physiol. 227, 981-986.

Scrutton, M. C. & Young, M. R. (1972) in The Enzymes (Boyer, P. D., ed.) vol. 6, 3rd edn, pp. 1-35, Academic Press, New York and London.

Utter, M. F. & Keech, D. B. (1960) J . Bid. Chem. 235, PC17. Utter, M. F. & Keech, D. B. (1963) J . Bid. Chem. 238, 2603-

Keech, D. B. & Utter, M. F. (1963) J . Bid. Chem. 238, 2609-

Wise, E. M., Jr & Ball, E. G. (1964) Proc. Nut1 Acad. Sci.

Mehlman, M. A,, Walter, P. & Lardy, H. A. (1967) J . Biol.

Ballad, F. J. & Hanson, R. W . (1967) J . Lipid Res. 8, 73-79. Melilman, M. A. (1968) J . Biol. Chem. 243, 3289-3295. Mahan,D. E., Mushahwar, I. K. & Koeppe, R. E. (1975) Bio-

Martin, B. R. & Denton, R. M. (1970) Biochem. J . 117, 861-

Mackall, J. C. & Lane, M. D. (1977) Biochem. Biophys. Re.?.

2608.

2614.

U.S.A. 52, 1255-1263.

Chem. 242,4594.

chem. J . 145,25-35.

877.

Commun. 79.720- 725.

and J. Nedergaard : Pyruvate Carboxylation in Brown Adipose Tissue

18. Martin, B. R. & Denton, R. M. (1971) Biochem. J . 125, 105-

19. Scrutton, M. C. (1978) FEBS Lett. 89, 1-9. 20. Cannon, B. & Nedergaard, J. (1978) in Effectors of Thermoge-

nesis (Girardier, L. & Seydoux, J., eds) Experientia Supple- mentum 32, pp. 107-111, Birkhauser Verlag, Base1 and Stuttgart.

21. Nedergaard, J . & Cannon, B. (1977) Ahsw. Commun. 11th Meet. Fed. Eur. Biochem. Soc. C-1 051.

22. Cannon, B., Nedergaard, J., Romert, L., Sundin, U. & Svar- tengren, J. (1978) in Strategies in Cold: Natural Torpidity and Thermogenesis (Wang, L. L. & Hudson, J., eds) pp. 567 - 594, Academic Press, New York.

23. Hittelman, K . J., Lindberg, 0. & Cannon. B. (1969) Eur. J . Biochem. 11, 183-192.

24. Pettersson, B. & Vallin, I. (1976) Eur. J . Riochem. 62, 383- 390.

25. Patel, M. S. & Hanson, R. W. (1970) J . Bid. Chem. 245,1302- 1310.

26. Gornall, A. G., Bardawill, C. W. & David, M. M. (1949) J . Bid. Chem. 177, 751 -766.

27. Cannon, B. & Lindberg, 0. (1978) in Methuds in Enzymology, Biomembrunes, Part D, Bioenergetics. I . Organelles and Mem- branes (Fleischer, S. & Packer, L., eds) chapter9, Academic Press, New York, in the press.

113.

28. Scrutton, M. C. (1974) J . Biol. Chem. 249, 7057-7067. 29. von Glutz, G. & Walter, P. (1976) FEBS L t ~ t . 72, 299-303. 30. Scrutton, M. C., Pearce, P. H. & Fatebene, F. (1977) Eur. J .

31. Cannon, B. (1971) Eur. J . Biochem. 23, 125-135. 32. Bernson, V. S. M. & Nicholls, D. G. (1974) Eur. J . Biochem.

33. Halestrap, A. P. & Denton, R. M. (1974) Biochem. J. 138,

34. Halestrap. A. P. (1975) Biochem. J . 148, 85 -96. 35. Halestrap, A. P. & Denton, R. M. (1975) Biochem. J . 148,

36. Parvin, R. & Pande, S. V. (3978) J . Bid. Chem. 253, 1944- 1946.

37. Pande, S. V. & Parvin, R. (1978) J . B i d Chem. 253, 1565- 1573.

38. Frieden, C. (1963) in The Enzymes (Boyer, P. D., Lardy, H. & Myrback, K., eds) vol. 7, pp. 3-24, Academic Press, New York and London.

39. Hull, D. & Hardman, M. J . (1970) in Brown Adipose Ti.we (Lindberg, O., ed.) pp. 97-115, Elsevier. New York.

40. Portet, R., Laury, M. C., Bertin, R., Senaiilt, C., Hluszko, M. T. & Chevillard, L. (1974) Proc. Soc. E q ? . Biol. Med. 147, 807-812.

41. Crabtree, B., Higgins, S. J. & Newsholme. E. A. (1972) Bio- chem. J . 130, 91 - 396.

42. Scrutton, M. C. (1975) Med. Hypothrses, 1. 207-213.

Biochem. 76, 219-231.

47, 517-525.

313-316.

97- 106.

B. Cannon and J. Nedergaard, Wenner-Grens Institut, Stockholms Universitet, Norrtullsgatan 16, S-11345 Stockholm, Sweden

the physiological role of pyruvate carboxylation in hamster brown adipose tissue

Documents