Hexose Metabolism in Pancreatic Islets. Ca2+-DEPENDENT ACTIVATION OF THE GLYCEROL PHOSPHATE SHUTTLE BY NUTRIENT. SECRETAGOGUES*.
Vol. 267, No. 19, Issue of July 5, pp. 13251-13256, 1992 Printed in U.S.A .
THEJOURNAL OF BIOLOGICAL CHEMISTRY
0 1992 by The American Society for Biochemistry and Molecular Biology, Inc
Hexose Metabolism in Pancreatic Islets Ca2+-DEPENDENT ACTIVATION OFT H E GLYCEROL PHOSPHATE SHUTTLEBY NUTRIENT SECRETAGOGUES* (Received for publication, August 22, 1991)
Abdullah Sener and Willy J. Malaisse From the Laboratory of Experimental Medicine, Brussels Free University, B-1070 Brussels, Belgium
mmol or 4325 k 243 cpm/pmol) were purchased from Du Pont-New A method is proposed for the measurement of the flux throughthe glycerol phosphateshuttle in pan- England Nuclear. The non-glucidic nutrient 2-ketoisocaproate (2creatic islets. Such aflux is taken as the ratio between KIC)' and the racemic mixture of the L-leucine analog 2-aminobicythe production of 3HOH and the specific radioactivity clo[2,2,l]heptane-2-carboxylate(BCH) were both obtained from Sigma. NAD-linked glycerol-3-phosphate dehydrogenase (EC 1.1.1.8) of ~-[2-~H]glycerophosphate in islets exposed to[2-3H] was provided by Boehringer (Mannheim, FRG). glycerol. D-Glucose and non-glucidic nutrientsecretaAll experiments were conducted in islets prepared from fed female gogues, such as 2-ketoisocaproate and 2-aminobicyalbino rats and incubated 37 at "C in a Krebs bicarbonate buffer (51, clo[2,2,l]heptane-2-carboxylate,stimulate, in a Ca2+- containing bovine serum albumin (1.0 mg/ml). In some experiments, dependent manner, circulation in the glycerol phos- the incubation medium contained no CaClz and was enriched with phate shuttle. The shuttle flux is commensurate with 0.5 mM EGTA. For the measurement of 3HOH production from [2-3H]glycerol, the fraction of pyruvate generation which is not coupled with L-lactate production. These findings support groups of 15-30 islets each were incubated in 30-40 p1 of medium. In the view that arise in D-glucose concentration leads tothese experiments, a tracer amount of [2-'H]glycerol was appropriactivation of the FAD-linked mitochondrial glycero- ately diluted with unlabeled glycerol. The incubation was halted by of 20 pl of a citrate-Na0H buffer (0.4 M, pH4.9) containing phosphate dehydrogenase through an increase in cy- 5addition mM KCN, 10 p M antimycin A, and 10 pM rotenone. The 3HOH tosolic Ca2+ concentration. formed during incubation was then measured as described elsewhere (6). After the 20-22-h incubation a t 20 "C required for the recovery of 3HOH, thesamples were frozen and kept a t -20 "C. They were then Inpancreatic islets, D-ghcosecauses a preferential and examined for their content in tritiated acidic metabolites. For this concentration-related stimulation of oxidative glycolysis, de- purpose, the samples were mixed with 0.2 ml of H20, sonicated (2 X fined as themodality of pyruvate formation notcoupled with 10 s) and applied on a ion-exchange chromatography column (AG1the generation of L-lactate (1).The preferential stimulation X8, formate form). A second 2-ml aliquot of H 2 0 was used to rinse tubes and also passedthrough thecolumn, which was then washed of oxidative, relative to total, glycolysis could be attributable the three timeswith 1.5 mlof H 2 0prior to elution with 1.0 M ammonium t o activation by Ca2+ of the FAD-linked mitochondrial gly- acetate, 0.1 M formic acid. The material eluted from the column is cerophosphate dehydrogenase, the activity of which is much herereferred toas acidicmetabolites. As inthe case of 3HOH higher in islets than other tissues (2-4). D-Glucose indeed production, all results were corrected for the blank value found in the increases the cytosolic concentration of Ca2+ in the B-cell, absence of islets. The islet content in unlabeled and tritiated L-glycerophosphate and Ca2+increases the affinityof the FAD-linkedenzyme for was measured in a separate series of experiments. Groups of 50-60 L-glycerophosphate (4). Inorder to fully validatesuch a islets each were incubated, as described above, in 30 p1 of incubation proposal,the flow throughthe glycerol phosphateshuttle medium containing, unlessotherwisementioned, 37 p~ of [2-'H] should be measured in intact islets. The present report pro- glycerol not mixed with any unlabeled glycerol. The incubation mevides the latter information, the shuttle flux being estimated dium thus contained 0.55 mCi/ml. After incubation, the medium was from the generationof 3HOH and the specific radioactivity of removed using a dissecting microscope and the islets immediately ~-[2-'H]glycerophosphate in islets exposed to [2-3H]glycerol. mixed with 140 p1 of HCl (50 mM), frozen in liquid N,, disrupted by This method is based on the knowledge that, in islet mito- mechanical vibration, and eventually heated for 10 min at 50-60 "C. a 2-min centrifugation at 5,000 X g, an aliquot (120 pl) of the chondria, equimolar amounts of 3HOH anddihydroxyacetone After HCl extract was neutralized to pH 7.0 by addition of 60 pl of NaOH phosphate are generated from ~-[2-~H]glycerophosphate (4) (0.1 M), the samples being then stored a t -20 "C. On the same day, so that the flux in the glycerol phosphate shuttle in intact external standards of either unlabeled L-glycerophosphate or ~ - [ 2 cells exposed to [2-3H]glycerol couldbejudged fromthe 'H]glycerophosphate (1.0 p ~ were ) prepared in HC1 (50 mM) and production of "OH after correction for the specific radioac- then processed in the same manner as theHCI islet extract. tivity of the precursor intracellular pool of ~-[2-~H]glycero- The assay of unlabeled or tritiated L-glycerophosphate was conducted on four aliquots (40 p1 each) of the neutralized HCI extract. phosphate. They were mixed with an equal volume of a Tris-HCI buffer (100 mM, pH 8.4) containing 1.0 mM EDTA, 20 mM ammonium acetate, MATERIALS AND METHODS 2.0 mM NADf, 2.84 p M 2-ketoglutarate, and 6.0 units/ml beef liver [2-'H]Glycerol (15Ci/mmol or 6494 _t 250 cpm/pmol), [U-14C] glutamate dehydrogenase. In the two samples used for the assay of unlabeled L-glycerophosphate, only [U-"C]2-ketoglutarate was presglycerol (342 mCi/mmol), and ~-[2-~H]glycerophosphate (10.6 Ci/ ent in the assay medium, as previously described (7). The samples * This work was supported by grants from the Belgian Foundation were incubated in the absence or presence of rabbit muscle glycerophosphate dehydrogenase (2.5 units/ml). After 60 min of incubation for Scientific Medical Research and Belgian State Prime Minister's Office-Science Policy Programming. The costs of publication of this article were defrayed in part by the payment of page charges. This The abbreviations used are: 2-KIC, 2-ketoisocaproate; BCH, 2article must therefore be hereby marked "advertisement" in accordaminobicyclo[2,2,l]heptane-2-carboxylate;EGTA, [ethylenebis(oxyance with 18 U.S.C. Section 1734 solely to indicate this fact. ethy1enenitrilo)ltetracetic acid; df,degrees of freedom.
13251
13252
Glycerol Phosphate Shuttle
at 20 "C, labeled L-glutamate was separated by ion-exchange chromatography as described elsewhere (7). The paired difference between the two samples was corrected for the blank reading attributable to glycerophosphate dehydrogenase and then taken asrepresentative of the islet content in L-glycerophosphate. The two other samples were used, in the same manner, for the assay of ~-[2-~H]glycerophosphate, in which case the assay medium did not containlabeled 2-ketoglutarate. The following features of the assay procedure should be emphasized. All results were calculated by reference to external standards treated in the same manner as the islet extract. In the assay of the unlabeled L-glycerophosphate standards, the blank reading (no substrate and no glycerophosphate dehydrogenase) represented 1.48 f 0.22% of the total radioactive content and was increased by 3.52 f 0.36% in the presence of glycerophosphate dehydrogenase ( n= 15 in both cases). The magnitude of the latter increase was related to both the concentration of glycerophosphate dehydrogenase and length of incubation. In theassay of tritiated ~-[2-~H]glycerophosphate standards, the reading recorded in the absence of glycerophosphate dehydrogenase represented only 0.15 f 0.03% (n = 6) of the total radioactivecontent. The NAD3H generated from ~-[2-~H]glycerophosphate and not recovered as tritiated L-glutamate failed to interfere with the measurement of the latter amino acid. Indeed when the samples were treated with a suspensionof charcoal to remove pyridine nucleotides (8) prior to chromatography, the results obtained with tritiated standards averaged 81.8 -+ 0.3% ( n = 4) of those recorded without charcoal treatment, the latter value being identical to the reading of 80.6 rt 0.6% ( n = 6) recorded after charcoal treatment of samples used for the assay on unlabeled L-glycerophosphate. [2-3H] Glycerol failed to react in this assay system. Likewise, in the absence of glycerophosphate dehydrogenase but presence of unlabeled Lglycerophosphate (25 pmol/sample), glutamate dehydrogenase failed t o act on the latter substrate. The generation of L-glutamate was grossly proportional to the amount of L-glycerophosphate present in the sample (Fig. 1). The efficiency of L-glutamate generation was close to 30% (Fig. 1).In this and other respects, there was no significant difference between the I I I J 0 generation of "C-labeled L-glutamate from unlabeled L-glycerophos10 20 30 40 phate, using [U-"C]2-ketoglutarate as thehydrogen acceptor, and the generation of tritiated L-glutamate from ~-[2-~H]glycerophosphate L-Glvcerophosphate (pmol/samplc) (Fig. 1). Although the interassay variability in the efficiency of the assay procedure was not negligible (Fig. l),the intraassay variability FIG. 1. Assay of unlabeledL-glycerophosphate (open ciras judged from the coefficient of variation for triplicate measurecles) and ~-[2-~H]glycerophosphate (closed circles) through ments, did not exceed 3.4 f 1.1%(assay of unlabeled L-glycerophos- the generation of I4C-labeled and tritiated L-glutamate, rephate) and3.0 f 1.2% (assay of ~-[2-~H]glycerophosphate), even with spectively. Mean values (* S.E.) are derived from six separate the lowest external standard (i.e. 10 pmol/sample). Relative to the experiments. corresponding external standards, internal standards(10 pmol/sample) yielded readings averaging 97.8 f 6.9% (in the case of unlabeled L-glycerophosphate) and 92.3 f 3.3% (assay of ~-[2-~H]glycerophos- ment, all available data in each set of experiments were eventually pooled together in the appropriate tables and figures. The standard phate). In a first series of experiments, the islets were extracted in 140 ~1 error on the ratio between mean values was calculated as described of perchloric acid (2.5%, v/v) instead of HCl, the external standards elsewhere (9). being also prepared in perchloric acid. After neutralization by addition RESULTS of 60 ~1 of a solution of KOH (1.0 M) and Tris (0.2 M), and treatment of the samples as described above, the efficiency of the assay proce[2-3H]Glycerol Conversion to 3HOH-The quantitatively dure was found to be about 3 times lower in the PCA rather than HCl extract, whether in the case of unlabeled or tritiated L-glycero- major tritiated metabolite generated from [2-3H]glycerol conphosphate. sisted in 3HOH. In the presence ofD-glUCOSe (16.7 mM), the The experiments dealing with the effect of non-glucidic nutrients concentration dependence for 3HOH production yielded an upon the metabolism of [2-3H]glycerolwere conducted in thepresence apparent K,,, close to 0.04 mM (Fig. 2). In the presence of 37 of2.8 mM D-glucose, because the islet content in both L - [ ~ - ~ H ] glycerophosphate and total L-glycerophosphate was much lower in p M glycerol, the production of 3HOH was grossly proportional the absence of the hexose than presence of 2.8 mMD-glUCOSe. As a to thelength of incubation (Fig. 3). A comparabletime course matter of fact, the concentration of unlabeled L-glycerophosphate in for 3HOH production was recorded in experimentsconducted islets deprived of exogenous nutrient for 120 min was close to the in the presence of either 4 p~ or 1.0 mM glycerol (data not limit of detection under the present experimental conditions. There- shown). A rise in D-glucose concentration from 2.8 to 16.7 fore, the specific radioactivity of tritiated L-glycerophosphate in the mM markedly increased the rate of 3HOH production from islets incubated in the absence of D-glucose could not be assessed in [2-3H]glycerol (Fig. 3). a reliable manner. In thepresence of 2.8 mM D-glucose, L-leucine (10 mM) also The production of"COZ from [U-'4C]glycerol was measured as augmented the production of 3HOH from [2-3H]glycerol (37 previously described (6). All results, including those already mentioned, are expressed as p M ) from 325 f 28 to 747 f 36 fm01/120 min/islet (n = 20 in the mean value (k S.E.) together with the number of individual both cases). Likewise, the deamination product of L-leucine, determinations (n)or degrees of freedom ( d f ) . The statistical signif- 2-ketoisocaproate, and the non-metabolized analog of L-leuicance of difference between mean values was assessed by use of (BCH), inStudent's t test. Such comparisons were restricted to data obtained cine, 2-aminobicyclo[2,2,1]heptane-2-carboxylate in close-to-equal number within the same experiments. However, in creased 3HOH production from [2-3H]glycerol (Table I). The order to avoid the cumbersome tabulation of each separate experi- absence of CaC12 and presence of EGTA (0.5 mM) failed to
cii
/
13253
Shuttle Glycerol Phosphate TABLEI
Effect of nutrient secretagogues and Ca" upon the generation of 3HOH from [Z-3Hlglycerol (37 p ~ ) 3HOH production (fmol/l20 min/ islet)
o-Glucose BCH 2-KIC mM
2932.8 16.7 2.8 2.8
mM
Ca2+(1.0 mM)
No Ca2'
f 13 268 (50) 618 f 22345 (60) 478 f 19 (40) f 373 (40) 16
f 16 (29) (19) f 19 328 f 18 (20) k 18 (19)
mM
10.0 490 10.0
TABLEI1 Effect of nutrient secretagogues and Ca2+ upon the generation of tritiated acidic metabolites from [2-3H]glycerol (37 phi) D-Glucose
2-KIC BCH
mM
mM
Tritiated acidic metabolites (fmol/ 120 min/islet) Ca2+(1.0 mM)
U
I OO
1/ 1
0.1
[
Glycerol] I
a2
2.8 16.7 2.8 2.8
(rnM") 013 '%o
No Ca2+
mM
10.0 10.0
33.1 f 2.8 (67) 34.9 f 2.6 (60) 35.8 f 2.4 (40) 38.0 f(39) 2.9
47.2 43.0 52.2 40.9
f 2.0 (53) f 2.3 (43) f 2.3 (20) f 1.8 (19)
impaired ( p < 0.001) in the Ca2+-deprivedmedium. Likewise, the absence of Ca2+significantly decreased 3HOH generation FIG. 2. Concentration dependence for 3HOH production from [2-3H]glycerol by islets exposed to either 2-ketoisocapfrom [2-3H]glycerol by islets incubated for 120 min in the roate or BCH in thepresence of 2.8 mM D-glucose (Table I). presence of 16.7 mM D-glucose. Mean values (fS.E.) refer to 9Generation of Tritiated Acidic Metabolites from [2-3H]Glyc33 individual measurements. addition to 3HOH, erol and Oxidation of [U-'4C]Glycerol-In the islets also generated tritiated acidic metabolites from [23Hlglycerol. After 120 min of incubation, the production of such tritiated acidic metabolites was about 1 order of magnitude lower than that of 3HOH. It was little affected by the concentration of nutrients and, asa rule, slightly increasedin the absence of Ca2+ (Table11). Its dependence on theconcentration of [2-3H]glycerol was similar to that found for the production of 3HOH, thevalues recorded in thepresence of 4 and 37 p~ averaging, respectively, 8.0 k 0.2 and 44.9 k 2.3% ( n = 18 in both cases) of that found in the presence of0.1 mM [2-3H]glycerol. The time course for the production of tritiated acidic metabolites differed, however, from that observed for the generation of 3HOH. Thus, after only 20 min of incubation in thepresence of 2.8 or 16.7 mM D-glucose, the net production of tritiated acidic metabolites already represented, respectively, 62.2 f 9.2 and 60.2 k 7.8% ( n = 30 in both cases) of the mean corresponding value recorded after 120 min of incubation. Although the generation from [2-3H]glycerolof radioactive metabolites other than 3HOH and tritiated acidic metabolites, e.g. glycerolipids, was not monitored in the present experiments, our results strongly suggest that a rise in D-glucose concentration augmented the phosphorylationof [2-'H]glycerol, at least at normal Ca2+concentration. In thepresence of Ca'+, a rise in D-glucose concentration from 2.8 to 16.7 mM also increased ( p aOOl) theoxidation of [U-'4C]glycerol T I M E (min) (Table 111). This effect was abolished in theabsence of extraFIG. 3. Time course for the production of 3HOH from [2-3H] cellular Ca2+.The oxidationof [U-'4C]glycerol and itsconversion of 14C-labeled acidic metabolites, when considered as a ) islets incubated in the presence of either glycerol (37 p ~ by 2.8 mM D-gluCOSe (closed circles and dashed line) or 16.7 mM whole, was not vastly different from the total production of D-glUCOSe (open circles and solid line). Mean values (fS.E.) 3HOH and tritiated acidic metabolites from [2-3H]glycerol refer to 24-39 individual measurements. (data not shown). Islet Content inTritiatedand Unlabeled L-Glycerophosaffect significantly the generation of 3HOH by islets exposed phate-To convert the production of 3HOH from exogenous t o 2.8 mM D-glucose. However, at a higher concentration of [2-3H]glycerol into a net flux through the glycerol phosphate the hexose (16.7 mM), the production of 3HOH was severely shuttle, the steady-state specific radioactivity of [2-'H]glyc[2-'1j Glycerol ( r n M )
Shuttle Glycerol Phosphate
13254
erol-3-phosphate in the islets needed to be known. For this (df, 28) of the control value recorded in the sole presence of 2.8 mM D-glucose. purpose, the islet content in both total ~-glycerol-3-phosphate A control set of experiments indicated that the specific and [2-3H]glycerol-3-phosphate was measured after 120 min of incubation in the presence of exogenous [2-3H]glycerol (37 radioactivity of tritiated L-glycerophosphate in islets after 120 p M ) . In these experiments, the exogenous tritiated substrate min of incubation reflected a close-to-steady-state situation. Indeed, when islets were incubated for either 60 or 120 min was not mixed with any unlabeled glycerol. Pooling all available measurements, the concentration of in thepresence of 16.7 mM D-glucose and [2-3H]glycerol (4 or , islet content in tritiated L-glycerophosphate and L-glycerophosphateaveraged 0.99 f 0.08 and 1.56 f 0.08 37 p ~ )the pmol/islet after 120 min of incubation in thepresence of 2.8 its specific radioactivity were not significantly affected by the and 16.7 mM D-glucose ( n = 70-71). When measured within length of incubation. Thus, after 120 min of incubation, the ~-[Z-~H]glycerophosphate andits specific the same experiments, thevalue recorded at the low concen- isletcontentin tration of D-glucose averaged 59.9 f 5.9% (df, 113) of that radioactivity averaged, respectively 112.2 k 18.6 and 80.2 f found at the high concentration of the hexose. At the low 12.4% (df, 16) of the correspondingvalues recorded after only glucose concentration, the incorporation of 2-ketoisocaproate 60 min of incubation. However, at the high concentration of or BCH (each 10 mM) into the medium also resulted in a D-glucose, there was a modest but significant increase in the modest increase of the islet contentin L-glycerophosphate to islet content of unlabeledL-glycerophosphatebetween the of incubation,themean earlyvalue respectively 122.5 f 10.3% and 133.2 f 9.1% (df, 28 in both 60thand120thmin averaging 71.4 -+ 7.2% (df, 18) of the mean late reading. The cases) of the control basal reading (Table IV). unlabeledL-glycerophosphate After 120 minof incubation in thepresence of 2.8 and 16.7 2.8 m ~ / 1 6 . 7 mM ratioin f 7.8% (df, 12)after only 60 min of amounted to 60.7 mMD-glUCOSe, together with [2-3H]glycerol (37 pM), the islet as such, was close to thatotherwise recorded content in ~-[2-~H]glycerophosphate did not exceed, respec- incubation and, after 120 min of incubation (see above). tively, 5.27 f 0.43 and 4.17 f 0.33 fmol/islet ( n = 45-49), The islet content in total L-glycerophosphate and tritiated when calculated by reference to the specific radioactivity of ~-[2-~H]glycerophosphate was also measuredafter 120 min of extracellular [2-3H]glycerol (37 pM). In the presence of 2.8 incubation in media deprived of extracellular CaC12 and conmM D-glucose, both 2-ketoisocaproate and BCH also tended taining 0.5 mM EGTA (Table IV). The absence of Ca2+ did t o lower the islet content in ~-[2-~H]glycerophosphate to, not prevent theincrease in theislet content of L-glycerophosrespectively, 85.5 f 7.6 and 80.8 f 7.6% (df, 28) of the control phate caused by a rise in D-glUCOSe concentration from 2.8 to basal reading. Incidentally, the ratio in ~-[2-~H]glycerophos-16.7 mM. The 2.8 m ~ / 1 6 . 7mM ratio in total L-glycerophosphate content of islets exposed to 4 p~ and 37 p~ [2-3H] phate indeed averaged, in the Ca2+-deprived media, 66.2 f glycerol averaged 27.1 f 6.9% (df, 18), in fair agreement with 6.8% (df, 34). However, in the absence of Ca2+, the rise in the ratio in 3HOH production as illustrated in Fig. 2. glucose concentration alsocaused an increase of the islet After 120 min of incubation in the presence of 2.8 mMDwith a 2.8 m ~ / 1 6 . 7mM content in ~-[2-~H]glycerophosphate, glucose and 37 p~ [2-3H]glycerol, the specific radioactivity of ratio averaging 66.0 f 6.5% (df, 34). As a result, the specific tritiated L-glycerophosphate averaged 7.29 f 0.64%0 (n = 46). radioactivity of tritiated L-glycerophosphate in islets deprived From datacollected within the same experiments, specific the of Ca2+ was no more significantly different at low and high radioactivity of tritiated L-glycerophosphate yielded a 16.7 glucose concentration. After exposure to 16.7 mM D-glucose m ~ / 2 . 8mM ratio of 43.2 f 9.4% (df, 63;p C 0.001 as compared in the Ca2+-deprived media, such a specific radioactivity avwith unity). Likewise, at the low hexose concentration, 2- eraged 141.1 f 7.6% ( n = 18) of the mean value recorded at ketoisocaproate and BCH lowered the specific radioactivity the samehigh D-glucose concentration in thepresence of Ca2+ of tritiated L-glycerophosphate to 69.0 f 8.5 and 60.2 f 8.5% (1.0 mM). The absence of extracellular Ca2+ also increased and its both the concentration of ~-[2-~H]glycerophosphate TABLE I11 specific radioactivity in islets exposed to 2-ketoisocaproate or Oxidation of [U-'4C]glycerol (37 p M ) BCH in thepresence of 2.8 m M D-glucose. Such was not the o-Glucose case, however, in the sole presence of 2.8 mM D-glucose. Ca2+ [U-"CJglyceroloxidation Flux in the Glycerol Phosphate Shuttle-To estimate the mM mM minlislet fmol/l20 rate of circulationin the glycerol phosphateshuttle,the 2.8 265 1.0 f(24) 22 418 f 35 (24) 1.0 16.7 production of 3HOH from exogenous [2-3H]glycerol was diNone 233 f 18 (24) 2.8 vided by the specific radioactivity of the precursor pool of None 262 k 22 (24) 16.7 (Fig. 4). A rise in D-glucose tritiated ~-[2-~H]glycerophosphate TABLEIV Effect of nutrient secretagogues and Ca2' upon the islet content in tritiated and total L-glycerophosphate after 120-min incubation in theDresence of 12-3Hlnlvcerol137 uM) o-Glucose
2-KIC
BCH
mM
mM
mM
CaZ+ mM
[2-'H]glycero-P
Glvcero-P
SDecific radioactivitv
% of control
% of control
% of control
n
100.0 f 3.9 1.0 100.0f 3.5 100.0 f 4.2 2.8" 43.2 f 3.3 204.2t 11.9 16.7 1.0 90.2f 4.2 69.0 f 5.4 85.5 f 5.4 122.5 k 8.5 1.0 2.8 10.0 133.2 f 6.9 60.2 t 5.4 80.8t 5.5 1.0 10.0 2.8 108.3 t 5.4 96.6 f 6.4 Nil 98.6f 5.0 2.8 78.3 f 4.2 Nil 134.1 f 6.0 170.6 f 8.8 16.7 96.0f 4.1 119.2 k 7.7 10.0 Nil 107.0f 5.3 2.8 117.1 f 6.0 110.3 f 6.0 92.2k 4.2 10.0 Nil 2.8 a The absolute control values (first line) averaged 5.27 f 0.43 fmol/islet for ~-[2-~H]glycerophosphate, 0.95 f 0.09 pmol/islet for total Lglycerophosphate, and 7.29 k 0.64 700 for the specific radioactivity of this metabolite.
Glycerol Phosphate Shuttle
13255
2-
c
FIG. 4. Metabolism of [2-3H]glycerol (37 BM) inisletsincubatedin the presence of either 2.8 mM Dglucose (open columns) or 16.7 mM D-ghCOSe (hatched columns).
9
"p
'0
I-
e
I
a -1 1-
TABLEV Effect of nutrient secretagogues and Cu2+upon circulation in the Plvcerol vhosvhute shuttle
tiated L-lactate). This interpretation is supported by both a prior study conducted in isolated hepatocytes exposed to [2*H]glycerol (11)and the finding that, after120 min of incuShuttle flux (pmol/l20 min/islet)" bation, comparable amounts of tritiated acidic metabolites o-Glucose 2-KIC BCH were recovered withinthe islets andinthe extracellular Ca" (1.0 mM) No Ca2+ incubation medium (data not shown). mM mM mM The present results also suggest that, at least in the pres40.2 2.8 38.1 (104) f 2.4 (60) f 3.4 ence of extracellular Ca2+, nutrient secretagogues increased 196.3 16.7 f 60.4 (92) 14.6 ? (35) 4.3 the phosphorylation rate of [2-3H]glycerol. By analogy with 2.8 10.0 95.0 k37.7 (53) 7.7 f(33) 3.2 2.8 10.0 111.6 f(53) 43.7 8.5 f(32) 3.1 the positive feedback regulation of D-glucose phosphorylation " Calculated from the data listed in Tables I and IV; mean values by endogenous ATP previously described in pancreatic islets (fS.E.) are given together with the corresponding degree of freedom (12), this may reflect the ATP dependence of glycerol phos(in parentheses). phorylation by glycerol kinase. Thus, in the islets as in hepatocytes, thisenzyme is bound, in part, tomitochondria (13, concentration from 2.8 to 16.7 mM caused a &fold increase in 14). Nutrientsecretagogues augment the rate of mitochondrial the shuttle flux. ATP generation and the latter effect is indeed impaired in As documented in Table V, both 2-ketoisocaproate and Ca2+-deprivedislets (6, 15). BCH also increased the circulation in theglycerol phosphate The nutrient-induced and Ca2+-dependent acceleration of shuttle, albeit to a lesser extent than observed at high D- glycerol phosphorylation may also account, in part at least, glucose concentration. In the absence of extracellular Ca2+, for the stimulationby D-glucose of 14C02production from [Uthe metabolic response to either 2-ketoisocaproate or BCH 14Clglycerol in islets incubated at normal extracellular Ca2+ was abolished. The response to a rise in D-glucose concentra- concentration. Moreover, insuch islets, the elevated D-glucose tion from 2.8 to 16.7 mM was also severely decreased, but not concentration may also favor, in a Ca2+-dependent manner, totally suppressed, in the absence of Ca2+. themitochondrialoxidation of [U-'4C]pyruvategenerated from [U-'4C]glycerol (15). In thisrespect, the presentfindings DISCUSSION which concern the fateof [U-'4C]glycerol tested at a very low concentration (37 p M ) differed from thosepreviously collected The present results confirm that, in pancreatic islets, Dglucose stimulates the production of 3HOH from [2-3H]glyc- inislets exposed to much higher concentrations (1.0-10.0 mM) of glycerol, in which case reciprocal competition between erol (10). They reveal that non-glucidic nutrientsecretagogues, such as L-leucine, 2-KIC, and BCH, reproduce the metabolites derived from the poly01 and from D-glucose might of 14C02(10). effect of D-glUCOSe and that the stimulant actionof both the interfere with the measured generation The rate of 3HOH production from exogenous [2-'H]glychexose and non-glucidic nutrients is inhibited in the absence erol must be divided by the specific radioactivity of the of Ca2+. to The islets also generated tritiated acidic metabolites from precursor pool of intracellular ~-[2-~H]glycerophosphate flow inthe glycerol phosphateshuttle. exogenous [2-3H]glycerol. The net productionof such metab- estimatethetrue olites was about 1 order of magnitude lower than that of Taking into account the concentrationdependence of [2-3H] metabolism,themeasurement of ~-[2-~H]glycero] "HOH and 1 order of magnitude higher than that of L - [ ~ - ~ H glycerol glycerophosphate. It was little affectedby the presence or phosphate, in samplesof the islet extract containing each the absence of nutrient secretagogues. This feature,as well as the material derived from about 10-11 islets, required the use of time course for the net productionof tritiated acidic metabo- [2-3H]glycerolin relatively high concentration (about 40 PM) lites, suggested thatitcorresponds,inpart,to a pool of withoutaddition of any unlabeled glycerol. Even so, the and intracellular metabolic intermediates (e.g. tritiated L-malate) radioactivity generated from ~-[2-~H]glycerophosphate rapidly reaching its steady-state value and, in part, to the recovered as tritiated L-glutamate was close to only 10' cpm/ time-related generation of metabolic end-product(s) (e.g. tri- sample. The specific radioactivity of islet cell ~-[2-'H]glycer-
13256
Shuttle Glycerol Phosphate
ophosphate indeed represented less than one percentof that of exogenous [2-3H]glycerol. In this respect, our enzymatic assay procedure allowed ustoidentifyin a most specific manner the tritium atoms in~-[2-~H]glycerophosphate. Our results clearly documentthat D-glucose andother nutrient secretagogues augment, ina Ca2+-dependent manner, the flux in the glycerol phosphate shuttle. The acceleration of the shuttle may thus be largely attributable to the activation of FAD-linked mitochondrial glycerophosphate dehydrogenase, as a result of the nutrient-inducedincrease in cytosolic Ca2+ concentration. In the case of the isletsexposed to a high concentration ofD-ghCOSe, the acceleration of the shuttle may also be due, to a limited extent, toa mass action phenomenonresulting from the hexose-inducedincrease inthe steady-state islet content of L-glycerophosphate. Thus,in such islets, the absence of extracellular Ca2+ severely decreased but failed to completely suppress the hexose-induced increment in either 3HOH production or shuttle flux. Theoretically, ourestimation of the glycerol phosphate shuttle flux could beslightlyoverestimated. Indeed, some in NAD3H could be generated from ~-[2-~H]glycerophosphate the reversible reaction catalyzed by the cytosolic NAD-dependent glycerophosphate dehydrogenase and then be converted to 3HOH in the mitochondria at the intervention of a shuttle other than the glycerol phosphateshuttle, e.g. the citrate-pyruvate shuttle (16). However, the postulated generation of cytosolic NAD3H would occur in a direction opposite to theprevailing cytosolic flux from glycerone-3-phosphateto L-glycerophosphate and NAD+. Moreover, whenever NAD3H is generated in the cytosol of islet cells, a sizable fraction of the tritiated nucleotide may be expected to be consumed in Only the conversion of unlabeled pyruvate to ~-[2-~H]lactate. the remaining fraction of NAD3H could lead to the mitochondrial generation of 3HOH, and,even so, this might be largely attributable to circulation in the glycerol phosphate shuttle. Since the generation of tritiated acidic metabolites from [2''Hlglycerol is 1 order of magnitude lower than that of 3HOH and since the production of ~ - [ 2 - ~ H ] l a c t aonly t e accounts for a fraction of the total generation of tritiated acidic metabolites (see above), the amount of NAD3H generated in the cytosol and susceptible to be transferred by a shuttle-mediated process into the mitochondriaobviously represents, at the best,a minor percentage of 3HOH production. As already mentioned, the glycerol phosphate shuttle mightitself largely participate in the mitochondrial transfer of this restricted cytosolic supply of NAD3H. In the lightof the present measurements, the generation of 3HOHnotattributabletooperation of the glycerol phosphate shuttle can be considered, therefore, as truly negligible relative to the totalproduction of 3HOH from [2-''H]glycerol.
Our measurements of the flow through the glycerol phosphate shuttle were close to, albeit possibly slightly higher, than those of oxidative glycolysis in the islets (l), provided that the latter is expressed as triose residues. Incidentally, the flow rate through the glycerol phosphate shuttle could conceivably exceed the rate of oxidative glycolysis, provided that reducing equivalents are simultaneously transferred from the mitochondria to the cytosol by another shuttle. For instance, suchcould be the case forthe malate-aspartate shuttle, which has been proposed to indeed export reducing equivalents from the mitochondria in pancreatic islets (17). The marked acceleration of the glycerol phosphate shuttle by non-glucidic nutrient secretagogues such as L-leucine, 2KIC, and BCH, and the pronounced influence thereupon of Ca2+ deprivation contrast with the quite modest effect of these environmental factors upon the rate of oxidative glycolysis in the islets (1).Conceivably, therefore, the changes causedby thesefactorsinthe flux throughthe glycerol phosphate shuttle could coincide with opposite modifications in the transferof reducing equivalents between mitochondria and cytosol, as possibly mediated by the malate-aspartate shuttle (17). In conclusion, the presentwork provides direct support for the concept that D-glUCOSe accelerates, in islet cells, circulation in theglycerol phosphate shuttle. Such aneffect appears attributable, in part a t least, to a Ca2+-induced activation of FAD-linked glycerophosphate dehydrogenase. Acknowledgments-We are grateful to J. Schoonheydt, M. Urbain, and G. Vandenbroeck for technical assistance and C. Demesmaeker for secretarial help. REFERENCES 1. Malaisse, W. J., Rasschaert, J., Conget, I., and Sener, A. (1991) Int. J. Biochem. 23,955-959 2. MacDonald, M. J. (1981) J. Biol. Chem. 256,8287-8290 3. MacDonald, M.J. (1982) Hormone Metab. Res. 14,678-679 4. Rasschaert, J., and Malaisse, W. J. (1991) Biochem. J. 278, 335-340 5. Malaisse-La ae, F., and Malaisse, W. J. (1984) in MethodsinDiabetes Research ?Lamer, J., and Pohl, S. L., eds) Vol. 1, pp. 147-152, John Wiley & Sons, New York 6. Malaisse, W. J., and Sener, A. (1988) Biochim. Biophys. Acta 9 7 1 , 246254 7. Sener, A,, and Malaisse, W. J. (1990) Anal. Biochem. 1 8 6 , 236-242 8. Lowry, 0.-H., and Passonneau, J. V. (1972)A Flexible System of Enzymatic Analysis, p. 141, Academic Press, New York 9. Sener, A,, Malaise-Lagae, F., Dufrane, S. P., and Malaisse, W. J. (1984) Biochem. J. 220,433-440 10. Sener, A,, Rasschaert, J., Zahner, D., and Malaisse, W. J. (1988) Int. J. Biochem. 20,595-598 11. Cronholm, T., and Curstedt, T. (1984) Biochem. J. 2 2 4 , 731-739 12. Carpinelli, A. R., Sener, A,, and Malaisse, W. J. (1987) Med. Sei. Res. 1 5 , 481-482 13. Ostlund, A,-K., Gohring, U., Krause, J., and Brdiczka, P. (1983) Biochem. Med. 30,231-245 14. Yilmaz, M. T.,Sener, A,, and Malaisse, W. J. (1987) Mol. Cell. Endocrinol. 52,251-256 15. Malaisse, W. J., and Sener, A. (1991) Mol. Cell. Biochem. 107, 95-102 16. Sener, A,, and Malaisse, W. J. (1991) Biochimie 7 3 , 1287-1290 17. Malaisse, W. J., Malaise-Lagae, F., and Sener, A. (1982) Endocrinology 1 1 1,392-397
