Quantifying the carboxylation of pyruvate in pancreatic islets.

Pyruvate has been estimated to enter the citric acid cycle in islets by carboxylation to the same extent or more than by decarboxylation. Those estimates were made assuming the dimethyl esters of [1,4-14C]succinate and [2-3-14C]succinate, incubated with islets at a concentration of 10 mM, gave the same ratio of 14CO yields as if [1-14C]acetate and [2-14C]acetate had been incubated. The labeled succinates, at 10 mM, but not 1 mM, are now shown to give ratios higher than the labeled acetates at those concentrations and therefore higher estimates when related to yields from [2-14C]glucose and [6-14C]glucose. Using the labeled acetate ratios in paired incubations, the rate of pyruvate carboxylation is still estimated to be about two-thirds the rate of pyruvate decarboxylation. Participation of the malic enzyme-catalyzed reaction explains the greater ratio of yields of 14CO from the succinates at 10 mM than 1 mM and increases in those ratios on glucose addition and can account for the removal from the citric acid cycle of oxaloacetate carbon formed in the carboxylation.

Pyruvate carboxylation to oxaloacetate on incubation of islets with glucose has been estimated to occur to as great an extent or more than pyruvate decarboxylation (1)(2)(3). If correct, that has important implications with regard to islet and presumably ␤ cell metabolism. Since in the turning of the citric acid cycle the amount of oxaloacetate is unchanged, i.e. oxaloacetate ϩ acetyl-CoA 3 citrate 3 2CO 2 ϩ oxaloacetate, at least half the carbon of glucose entering the citric acid cycle would have to leave the cycle in a product or products other than CO 2 . The formation of so much oxaloacetate and hence its product(s) would presumably have a purpose.
The estimates depend in principle upon a comparison of the ratio of the yields of 14 [2-14 C]acetate, respectively, if pyruvate's metabolism were only via decarboxylation to acetyl-CoA, the ratios from the labeled pyruvates and acetates should be the same (Fig. 1). Since [1-14 C]acetate is oxidized to 14 CO 2 in fewer turns of the citric acid cycle than [2-14 C]acetate, unless the only fate of the carbons of acetate is to CO 2 , the yield of 14 CO 2 from [1-14 C]acetate will exceed that from [2-14 C]acetate. If pyruvate is carboxylated because of rapid equilibration be-tween oxaloacetate and fumarate, both labeled pyruvates should yield [2,[3][4][5][6][7][8][9][10][11][12][13][14] C]oxaloacetate (Fig. 1). If equilibration is complete, the yield of 14 CO 2 from both pyruvates would be the same. Therefore, to the extent there is carboxylation relative to decarboxylation, the ratio of the yields from the labeled acetates should exceed that from the labeled pyruvates.
MacDonald (2,3) (4). Actually, the dimethyl esters of the labeled succinates were used, since the esters penetrate the cell membrane and once inside the cell are hydrolyzed. Ratios of yields of 14 CO 2 from the labeled glucoses were about 2. Ratios of 4 -6 or more, the larger in the presence of unlabeled glucose, were observed when the labeled succinates were incubated at 10 mM concentration, resulting in the quantitation of equal or greater rates of pyruvate carboxylation than decarboxylation.
Uncertainty exists with regard to the quantitation. First, the method of Kelleher and Bryan (4) specifies the use of the paired flask technique, i.e. labeled pyruvate in the presence of unlabeled acetate or equivalent and labeled acetate in the presence of unlabeled pyruvate or equivalent incubated under identical conditions. Second, the metabolism of acetate and succinate are not equivalent, even though the fate of acetate and succinate tracers will be identical if incubated under identical conditions. In the citric acid cycle a quantity of acetate, but not succinate, is oxidized to CO 2 , i.e. succinate 3 oxaloacetate ϩ acetyl-CoA 3 citrate 3 2CO 2 ϩ succinate. We have now incubated the 14 C-labeled glucoses, acetates and dimethyl succinates with islets using the paired flask technique and from the ratios of the yields of 14 CO 2 quantified pyruvate carboxylation. Islets-Islets were isolated from pancreata of fed 200 -300-g male Sprague-Dawley rats. The pancreata were digested with collagenase, and islets in the digest were collected under stereomicroscopy using a glass pipette. Unless otherwise noted, the islets were cultured for 20 h in RPMI 1640 medium, 11 mM glucose, 10% fetal calf serum, 100 units of pencillin G/ml, and 100 g of streptomycin/ml. They were preincu-bated for 30 min in 3.3 mM glucose and then washed three times with Hanks' solution before use.

Materials-D-[2-
Incubations-In the first series of experiments, in each experiment islets were distributed sequentially in cups of 1-ml volume to a total of 50 islets/cup. The cups hung from rubber stoppers closing 20-ml vials, and each cup contained 0.1 ml of Krebs-Ringer bicarbonate. In one cup [1,4-14 C]succinate and in another [2,3-14 C]succinate were at 1 mM concentration. In a third cup [1,4- C]glucose at 20 mM. In a second series, the contents of the cups were identical except that in the first six cups there was also unlabeled glucose at 20 mM and in the remaining two unlabeled acetate at 1 mM. Each cup had 1.1 Ci of 14 C in the labeled compound added, except for only 0.8 Ci of [2,3-14 C]succinate. In a third series one cup contained [1-14 C]acetate and the other [2-14 C]acetate, at 1 mM. Two other cups were the same except that the acetate was at 10 mM, and the last two cups also had labeled acetate at 10 mM but in the presence of 20 mM unlabeled glucose. In a fourth series labeled acetate was incubated with unlabeled glucose and unlabeled acetate with labeled glucose, but the islets were fresh, i.e. not cultured. Incubates were also prepared with identical contents except no islets were added. Vials with their contents were gassed with 95%-5% CO 2 for 2 min and then incubated at 37°C for 90 min while being shaken.
Measurement-0.1 ml of 10% perchloric acid was injected into each cup and 1.5 ml of CO 2 -free 1 N NaOH and 0.5 ml of 1 N NaHCO 3 placed by injection at the bottom of the vial. The vials with their contents then were kept at 37°C for 2 h to absorb into the NaOH the CO 2 evolved on addition of the acid.
The cups were removed and 2 ml of 5% BaCl 2 were added to each vial. The barium carbonate that precipitated was collected by filtering under suction the contents of the vial onto a preweighed filter paper. The barium carbonate that collected on the paper was washed with CO 2 -free water, dried and weighed. The barium carbonates weighed between 96 and 110 mg, about the theoretical yield from 0.5 mmol of NaHCO 3 .
The barium carbonate, still on filter paper, was placed at the bottom of a wide-mouth bottle containing 5 ml of water and closed with a rubber stopper from which a scintillation vial containing 2 ml of Hyamine was suspended. After evacuating air from the bottle through the stopper, 2 ml of 1 N H 2 SO 4 was injected through the stopper into the water. The bottle with its contents was kept at 37°C for 2 h to allow the CO 2 evolved from the barium carbonate to be absorbed into the Hyamine. Scintillation fluid was then added and 14 C activity assayed in a scintillation counter.
Calculations-Yields of 14 CO 2 in the incubates in the absence of islets were only 0.002-0.003% of the added 14 C for all the labeled compounds, except 0.02% for [1,4-14 C]succinate. Yields of 14 CO 2 in disintegrations/min were calculated by subtracting the relatively small number of disintegrations/min in CO 2 collected in the absence of islets from the disintegrations/min in their presence. At concentrations of 1, 10, and 20 mM the quantities incubated were 100, 1000, and 2000 nmol, respectively. Yields of 14 CO 2 expressed in nanomoles to 14 CO 2 /50 islets/90 min of incubation were calculated by multiplying the quantities incubated by the yields of 14 CO 2 in disintegrations/min and dividing by the disintegrations/min incubated. From the ratio of the yields from [1,4- where A 14 CO 2 is the ratio from labeled acetate or succinate, Pyr 14 CO 2 is the ratio from labeled glucose, and F, assumed to be 0.8, the ratio of 14 C from [6-14 C]glucose, the equivalent of [3-14 C]pyruvate, in carbon 2 to carbon 3 of oxaloacetate (1,2).
Statistical Analyses-Ratios and yields are recorded as mean Ϯ S.E. The significance of differences between the ratios was assessed using Student's t test for unpaired observations.

RESULTS
Yields of 14 CO 2 and the ratios of those yields in the first two series of experiments are recorded in Table I. The ratio of the 14 CO 2 yield from [2-14 C]glucose to that from [6-14 C]glucose at a glucose concentration of 20 mM was the same in the absence and presence of 1 mM acetate (means of 1.55 and 1.58). The ratios of the yields at 1 mM acetate were also the same in the absence and presence of 20 mM glucose (means of 2.64 and 2.72) and significantly more, p Ͻ 0.05 and p Ͻ 0.001, respectively, than the glucose ratios (1.55 and 1.58). The ratios of yields from the succinates were also significantly more than those from the labeled glucoses (p Ͻ 0.001, except p Ͻ 0.05 for 2.35 versus 1.55). The ratio at 1 mM succinate was not different from the ratio at 1 mM acetate in the absence (2.35 versus 2.64) but was in the presence of glucose (3.46 versus 2.72, p Ͻ 0.05). The ratios for succinate at 1 and 10 mM succinate were more in the presence than absence of 20 mM glucose (3.46 versus 2.35, p Ͻ 0.01, and 4.47 versus 3.05, p Ͻ 0.025). At 10 mM succinate in the presence of glucose the ratio was significantly more than for acetate (4.47 compared with 2.72, p Ͻ 0.01). At 10 mM succinate 3-5 times as much succinate was oxidized to CO 2 as at 1 mM succinate. The increase in the ratio of the yields from succinate on unlabeled glucose addition was due mostly to a decrease in the yield of 14 CO 2 from [2,3-14 C]succinate.
The ratios of 1.55 for Pyr 14 CO 2 and 2.64 for A 14 CO 2 calculate to a pyruvate carboxylation contribution of 43%. Using 2.35 for A 14 CO 2 the contribution is 39% and using 3.05 it is 48%. The ratios of 1.58 for Pyr 14 CO 2 and 2.72 for A 14 CO 2 give a contribution of 43%; with 3.46 it is 49%, and with 4.47 it is 54%.
In the series of experiments comparing the ratios of yields at 1 and 10 mM acetate, and 10 mM acetate in the presence of glucose, there were no significant differences among the ratios (Table II). The ratio at 1 mM acetate in that series, 2.62 Ϯ 0.24, was the same as in the first series, 2.64 Ϯ 0.33 (Table I). In contrast to succinate, increasing acetate's concentration 10-fold resulted in less than a 2-fold increase in its oxidation to CO 2 . Furthermore, the addition of glucose had no effect on the yield of 14 CO 2 . Incubations of fresh islets with 1 mM acetate and 20 mM glucose (Table III)   moved when the BaCO 3 was collected. The conditions of islet preparation and incubation we used were similar to those used by MacDonald (2,3). Islets were cultured by him in medium containing 1, 5, 8, and 20 mM glucose rather than just 11 mM glucose.
Succinate used in a trace quantity, so that metabolism is not altered, should in theory be able to replace acetate. That is evidenced from similar ratios at 1 mM succinate and acetate. However, the addition of succinate as a substrate can alter metabolism. That is evidenced in the present study by 1) similar ratios at 1 and 10 mM acetate but not 1 and 10 mM succinate, 2) the greater increase in succinate's oxidation to CO 2 with the increase in concentration, and 3) the increase with glucose addition of the ratios with labeled succinate but not acetate.
Succinate is converted to oxaloacetate in the citric acid cycle (Fig. 2), and in one turn of the cycle succinate is regenerated, albeit containing the carbons of the acetyl-CoA that condensed with the oxaloacetate. Therefore, when succinate enters the cycle in substrate amounts, at steady state, the quantity of carbon leaving the cycle must equal the quantity in the succinate entering, and that amount cannot leave as CO 2 . To the extent succinate's uptake exceeds the amount of acetyl-CoA available for condensation, that succinate can only be metabolized in a segment of the cycle. That segment, the dicarboxylic acid segment, is from succinate to oxaloacetate, and the amount of carbon leaving the segment, equal to the amount in the succinate taken up, can only be as malate and/or aspartate (oxaloacetate and fumarate do not cross the inner mitochondrial membrane).
Following the transport of malate from the mitochondrial matrix to the cytosol, [1,4- [2,[3][4][5][6][7][8][9][10][11][12][13][14] C]pyruvate and no 14 CO 2 . Thus, the greater ratio of 14 CO 2 yields, coupled with the 3-5-fold greater oxidation of the succinate to CO 2 at 10 mM than at 1 mM, with no source of exogenous acetyl-CoA, is evidence for much of the succinate at 10 mM being metabolized via the malic enzymecatalyzed pathway. The increased ratio on glucose addition can be explained by dilution of labeled pyruvate by unlabeled pyruvate formed from the glucose. As a result, a smaller amount of labeled pyruvate formed from labeled succinate would be expected to be oxidized. The greater reduction in the yield of 14 CO 2 from [2,3-14 C]succinate than from [1,4-14 C]succinate (Table I)    nate but not [2,[3][4][5][6][7][8][9][10][11][12][13][14] MacDonald demonstrated the presence of malic enzyme in islets (1,5). Malaisse et al. (6), incubating islets with glutamine at a substrate concentration, 10 mM, reported that a major fraction of the glutamine, via conversion to ␣-ketoglutarate, left the cycle as malate, which was converted to pyruvate. However, Malaisse and Sener (7)  Randomization of carbon 2 of glucose-6-P in the pentose cycle is assumed not to affect the ratio of the 14 CO 2 yields. The relative small contribution of the pentose cycle to glucose utilization by islets supports that assumption (2,7). If there is significant conversion of dihydroxyacetone-3-P to glycerol or its derivatives, isotopic equilibration of the dihydroxyacetone-3-P with glyceraldehyde-3-P is assumed sufficiently complete so as not to result in an overestimation of carboxylation. Relative high activity of triose-P isomerase in islets (8), a relatively small incorporation of glucose carbon into lipid via glycerol 3-phosphate (9), and only slightly less yields of 14  The assumption that F ϭ 0.8 in the calculations, i.e. that there is extensive equilibration between oxaloacetate and fumarate, is supported by high activities of malic dehydrogenase and fumarase in islet mitochondria (2) and incorporations of about 80% as much 14 C from [3-14 C]lactate into carbons 2 and 5 as carbons 1 and 6 of glucose formed by liver and kidney (10,11). However, in the presence of a pool of unlabeled succinate formed from dimethyl succinate and a pool of labeled oxaloacetate formed from labeled pyruvate, isotopic equilibration of the dicarboxylic acids should be less than in the absence of the unlabeled succinate pool. An estimate for example of 54%, assuming F ϭ 0.8, decreases to 32% if there is no equilibration, F ϭ 0. An assumption of course is also that there are single pools of intermediates, e.g. acetate enters the same pool of acetyl-CoA as acetyl-CoA formed from pyruvate. That goes beyond any concern that the islet contains several cell types, even though ␤ cells predominate.
In conclusion, the quantitation of pyruvate carboxylation in islets has been examined. It is estimated to proceed at about two-thirds the rate of pyruvate decarboxylation. The malate enzyme-catalyzed reaction allows for the removal from the citric acid cycle of the oxaloacetate formed. That is in accord with the recent report that when mitochondria from islets were incubated with [1-14 C]pyruvate, 14 C was recovered in the incubation medium mainly in malate, and when islets were incubated with [U-14 C]succinate 14 C appeared in pyruvate and lactate (12). Cytosolic NADPH would then be generated for cellular needs. There would then be cycling as the oxaloacetate, formed by fixation of pyruvate by CO 2 , is decarboxylated via malate and pyruvate is reformed (Fig. 3). A portion of the pyruvate would then be recarboxylated to oxaloacetate with the remainder decarboxylated to acetyl-CoA and CO 2 or reduced to lactate.