Insulin Regulation of Branched Chain a-Keto Acid Dehydrogenase in ...

2 downloads 0 Views 757KB Size Report
gassed with 95% O,, 5% COi, and placed in a shaking water bath. (37°C) for 30 .... After the indicated interval, each aliquot was chilled in an ice bath, stored at ...
JOURNALO F B I O L O G I C A L CHEMISTRY Vol. 255. No. 13. Issue of July IO. pp. 61%-6192, 1980 Prrnfed zn I: S A .

'rHE

Insulin Regulationof Branched Chaina-Keto Acid Dehydrogenase in Adipose Tissue* (Heceived for publication, November 8. 1979)

G. Peter Frick and H. Maurice Goodman From the Departmentof Physiology, Unicersity of Massachusetts Medical School, Worcester, MassachusettsOlfiOFi

The enzyme which oxidizes a-keto[ l-14C]isocaproate increase the transportof amino acids acrossthe plasma memto I4CO2 is activated by incubation of adipose tissue branes of adipocytes (21-23) or accelerate its transamination segments with insulin. A %foldreduction in the appar- toa-ketoisocaproate (20). Thesestudies provided indirect ent K , of the enzyme for a-ketoisocaproate was ob- evidence that decarboxylation of a-ketoisocaproate to form served when homogenates of adipose tissue segments isovaleryl-CoA may be both rate-limiting and the reaction treated with insulin were compared to homogenates of accelerated by insulin (20). It was found that while insulin control tissues. The enzyme was assayed at various stimulated the production of I4CO2in adiposetissuewhen times after homogenization of adipose tissue segments. either L-[l-'4CJleucine or a-ket~[l-'~C]isocaproate was used Relatively small changes were observed in the activity as substrate, it failed to increase the rate of utilization of [lfrom control or insulin-treated tissues for 30 min after "C]isovalerate. homogenization. The persistence of the insulin effect Branched chain a-keto acid dehydrogenase has been puriafter homogenization suggests that insulin may cause fied from extractsof liver and kidney mitochondria (11-13). It a covalent modification of the enzyme. The possibility appears to be a multienzyme complex similar to the pyruvate that a-ketoisocaproate is oxidized by pyruvate dehydrogenase, which is also stimulated by insulin, is un- dehydrogenase complex. In adipose tissue, the activation of branched chain a-keto acid dehydrogenase might be analolikely since the enzyme responsible for oxidation of I 4 C labeled branched chain a-keto acids can be inactivated gous to thewell known action of insulin on pyruvate dehydroby heat at a rate distinct from that of pyruvate dehy- genase (24-29). Branched chain a-keto acid dehydrogenase drogenase. Moreover, unlabeled branched chain a-keto has not been demonstrated previously in adipose tissue (14). acids inhibit the oxidation of a-ket~[l-'~C]isocaproate By using a-keto[ l-''C]isocaproatewith a highspecific activity, but not that of [l-L4C]pyruvate.Branched chain a-keto we have employed an assay sufficiently sensitive to study the acid dehydrogenase can be activated by incubation of effect of insulin on the enzyme in adipose tissue extracts. Our adipose tissue homogenates in the presence of magne- experiments have been modeled after those used to demonsium chloride and in the absence of ATP. The addition strate effects of insulin on pyruvate dehydrogenasein adipose of ATP plus an ATP-regenerating system reverses the tissue (24-29). activation of the enzyme. The apparent K , of the enEXPERIMENTAL PROCEDURES zyme is reduced and theV,,, is increased by incubation of tissue extracts under appropriate conditions. Male rats weighing 150 to 250 g were obtained from Charles River

The branched chain amino acids, leucine, isoleucine, and valine, appear to form a distinct class of amino acids whose degradation occurs primarilyat extrahepatic sites (2-8). These amino acids appear to share enzymes needed both for their initial transamination ( 3 , 9 , 10) and for the subsequent decarboxylation of their a-ketoanalogs (11-14). Leucine appears to be the most extensively studied of the branched chain amino acids. Muscle, because of its mass and its abundanceof transaminating enzymes, seemsto be the most importantsite quantitatively for leucine degradation (IS-19). Intherat, adipose tissue, which readily converts the carbon skeletonof leucine to CO, and lipids (4,8, 15-20), has been estimated to be the second most important tissue (8). Studies in this (17, 20) and other laboratories (4,8, 18, 19) established that insulin promotes the degradation of leucine in adipose tissue, but the mannerin which it acts has not been fully elaborated. Earlier studiessuggested that insulin did not * These studies were supported bv LJnited States Public Health ServiceGrant AM21216 and by PostdoctoralFellowshipAward AM05542 (G. P. F.). l'reliminarvreports of this work havebeen published (1,2). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore beherebymarked"ctdwrtisuntmt"inaccordancewith18 (J.S.C. Section 1734 solely to indicate this fact.

Breeding Laboratories (CD strain) and fed Purina Formulab Chow 5008 ad libitum. Kats were killed by cervical dislocation and the thin portion of the epididymal fat pad was rapidly excised. Adipose tissue segments weighing 100 to 300 mg were incubated in 1 mlof KrebsRinger bicarbonate buffer (30) containing 1.0 mM CaCli and 11 mM fructose.Incubationswerecarriedoutinsealedvialswhichwere gassedwith 95% O,, 5% COi, and placed in a shaking water bath (37°C) for 30 min. Tissues were treated with insulin by transferring them to fresh buffer containing 1 milliunit/ml insulin (Lillv, Iletin) for a second incubation lasting60 min. Tissues were homogenized at 0°C in a volume of buffer ( I O mM potassium phosphate, pH 7.4, plus 1 mM dithioervthritol) equal to 4 times the tissue weight. using a hand operated Ten Broeck homogenizer or a motor-driven Dual1 homogenizer. T h e homogenate was filtered througha plug of glass wool in a Pasteur pipetteor centrifuged at 'LOO X g for 30 s, and the filtrate or infranate was stored ice on until it was assayed for pyruvate dehydrogenase or branched chain ct-keto acid dehydrogenase. Assay for tr-Keto Acid L)ehydrt)~enases-Branched chain tr-keto acid dehydrogenase was assayed bv measurement of the release of I8 COLfromn-keto[]-"Clisocaproate, essentially as described by Wohl10 mM heuterandHarper(14).Thereactionmixturecontained: sodiumbicarbonate, 10 mM potassiumphosphate, pH 7.4, 1 mM BIITA, i mM dithioerythritol. 3 IIIM magnesium chloride, 0.15 mM coenzyme A, 1.5 mM NAD, 0.10 n m thiaminepyrophosphate. 1'; bovineserumalbumin(Sigma,essentiallvfattyacid-free,dialvzed against 0.1 M Tris-HC1, pH 7.0. and 10 mM EDTA. followed by several changes of distilled water), and the indicatedconcentration o f oketo[l-"Clisocaproate(approximately 2 pCi/pmol).Background "CO, was assessed bv analysis of assavs in which homogenization

6186

7

in Adipose Tissue

Branched Chain a-Keto Acid Dehydrogenase

6187

Because the observed rate of n-ketoisocaproate oxidation buffer replaced the tissue extract. Background "COL is probably due to nonenzymatic breakdown of the substrate; the procedures de- was small compared to that of pylL"ate, we examined the scribed consistently resulted in background I4CO2equal to 0.3 to 0.5% possibility that branched chaina-keto a c L might be oxidized of the total I4C added. slowly by pyruvate dehydrogenase rather than by a separate or a-keto[U-"C]-P-methyI-nOxidation of a-ket~[l-'~C]isovalerate enzyme. Extracts of adipose tissue were partially inactivated valerate was measured by substituting the indicated compound (2 or 13 gCi/pmol, respectively) for a-keto[ l-'4C]isocaproate. Pyruvate de- by heating them to 44°C for 0 to 30 min. The extracts were for branched hydrogenase was assayed as described for branched chain a-keto acid then assayed for pyruvate dehydrogenase or of dehydrogenase except that approximately 0.6 mM [l-'4C]pyruvate chain a-keto acid dehydrogenase (Fig. 1). The oxidation (0.5 pCi/pmol) was substituted for a-ket~[l-'~CC]isocaproate, and the four substrates,a-keto[ l-'4C]isocaproate, a-keto[ l-'4C]isovalreaction was terminated after 2 min. a-Keto[l-"Clisocaproate was erate, a-ket~[U-'~C]-P-methyl-rl valerate, and [ l-"C]pyruused within 1 week after being dissolved in 100 pl of 0.1 N HCI. The vate, was measured. The activityof p:,~..x7ate dehydrogenase dissolved substrate was kept frozen or ice-cold and was added to the other components of the reaction mixture just before the assay. declined at a slower rate t h a n that of the enzyme (orenzymes) Assays were initiated either by addition of tissue extract (200 to 300 responsible for oxidation of the branched chain a-keto acids. p l ) to the other components of the reaction mixture or by addition of After 30 min the pyruvate dehydrogenase activity declined to 50 p1 of substrate. Vials were placed in a shaking water bath (37°C) 36% of its initial value, but the branched chain a-keto acid 1 min before initiating the assay. Assays were routinely performed in dehydrogenase declined to only 7.8% of its initial value. The triplicate, using vials sealed with serum stoppers from which plastic oxidation rates of all three branched chain a-keto acids decups were suspended inside the vial by a stainless steel wire. Reactions were terminated after 5 min by injection of 0.25 ml of 1 N HASOIinto clined from the initial values at approximately the same rate, the reaction mixture. One minute prior to injecting the acid, 0.25 ml suggesting that a single enzyme may be responsible for their of phenethylamine was injected into the plastic cup. The vials were oxidation. When the heat inactivation was conducted at a shaken for 30 min to trap "CO,, whichwas quantitated by liquid highertemperature (52"C), the resultswerequitesimilar, scintillation spectrometry. Enzyme activity is expressed in milliunits; except that the enzymes were inactivated much more quickly; 1 milliunit is defined as the amount of enzyme required to catalyze after 2 min, the pyruvate dehydrogenase activity declined to the formation of 1 nmol of "COZ/min at 37°C. Synthesis ofBranched Chain a-Keto Acids-L-Amino acid oxidase 47% of its initial value and the branched chain a-keto acid 24% of its inital value was used to convert "C-labeled L-amino acids to a-keto acids by a dehydrogenaseactivitydeclinedto procedure based on that described by Odessey (31). ~-[l-'~C]Leucine (results not shown). (250 pCi) in2% ethanol was evaporated to dryness and redissolved in For the substrates [ l-I4C]pyruvate, a-keto[ 1-"Clisocap0.5mlof 50 mM Tris-HC1, pH 7.8, containing sufficient unlabeled roate, and a-ket~[l-'~C]isovaleratethere is noambiguity leucine to yield the desired specific activity. Catalase (20,000 units) and L-amino acid oxidase (2 units) were added, and the mixture was 0 incubated at 37°Cfor 2 h. The reaction mixture was applied as a 0 streak to a thin layer chromatography plate coated with Silica Gel G. -01 The chromatogram was developed in the solvent: chloroform/amyl -02acetate/formic acid (l:l:l, by volume). The chromatogram was dried for 60 min in a gentle stream of air. The product was located by -03autoradiography after exposure of the x-ray film for 2 h. Little or no -04 ~-[l-'~CC]leucine was observed. The silica gel containing the labeled compound was scraped off and extracted with 0.01 N NaOH (8 ml). -05The extract was passed through a small ion exchange resin (Bio-Rad -. - 0 6 AG 50-X8,0.4-cmdiameter X 1-cm long,H' form) to remove calcium, \ = neutralized, lyophilized, and stored at -20°C. The identity and speA- -07were confirmed by formation cific activity of a-ket~[l-'~C]isocaproate Ln of the 2,4-dinitrophenylhydrazoneand measurement of the absorb_o -08ance at 440 nm as described by Taylor and Jenkins (32). In addition, similar results were obtained using substrate prepared from leucine oflow specific activity or from leucine of high specific activity to which unlabeled a-ketoisocaproate was added. a-Ket~[U-'~C]-P-methyl-n-valerate was prepared from L-[U-'T]isoleucine and a-ket~[l-'~C]isovaleratewas prepared from ~ - [ l -I z { "CC]valine by the procedure above. Oxidation of the amino acids was - I 34 O i terminated by addition of HCI (final concentration 1 N), and the reaction mixtures were passed through ion exchange columns as 6 1 8 24 30 described above, adjusted to pH 6 with potassium phosphate buffer, Tlme, mtn and lyophilized. Analysis of the products by thin layer chromatography and autoradiography revealed only the branched chain a-keto FIG. 1. Heat inactivation of pyruvate dehydrogenase and acids. branched chain a-keto acid dehydrogenase. Epididymal fat was homogenizedinice-coldbuffer as described under "Experimental RESULTS Procedures," and the infranate was recovered after centrifugation. An enzyme capableof oxidizing branched chain a-keto acids Magnesium chloride was added to make the concentration 5 mM, and the extract was incubated at 37°C for 30 min. Aliquots (4 ml) were was found inadipose tissueand was assayed bymeasuring the transferred to 25-1111flasks equilibrated in a 44'C shaking water bath. release of I4CO, from a-keto[l-'4C]isocaproate. Pyruvate deAfter the indicated interval, each aliquot was chilled in an ice bath, hydrogenase was assayed using a similar reaction mixture stored at 0°C for up to 40 min, and assayed for pyruvate dehydrogencontaining [ l-14C]pyruvate instead of a-keto[ l-'4C]isocapase or branched chain a-keto acid dehydrogenase in a reaction mixture roate. The oxidation of both a-ketoisocaproate and pyruvate containing a I4C-labeled substrate at the following concentration: was stimulated by addition of coenzyme A, NAD, and thia; 200 p ~ a-ketoisovalerate, ; 200 pyruvate, 600 p ~ a-ketoisocaproate, mine pyrophosphate to adiposetissue extracts. In the absence p ~ or ; a-keto-/3-methyl-n-valerate, 25 p ~ The . rates of oxidation observed before heat inactivation were (milliunits per g of tissue): of added cofactors, the rate of oxidation of a-ketoisocaproate pyruvate, 243; a-ketoisocaproate, 6.23; a-ketoisovalerate, 11.1; and awas approximately10%of the rate observed withthe complete keto-P-methyl-n-valerate, 3.57. The log of the ratio of the activity a t reaction mixture. The residual a-ketoisocaproate oxidation time t (u,) to theinitial activity (u,) is plotted as a function of the time present in the interval at 44OC. The curves were fitted by the methods of least may be attributed to endogenous cofactors tissue extract since the rate of oxidation of pyruvate was squares to the results obtained with pyruvate and to the results reduced to a similar extent by omission of the cofactors. obtained with all three branched chain a-keto acids.

:\

1

\

6

,;

Branched Chain a-Keto Acid Dehydrogenase in Adipose Tissue

6 188

Q-Keto [ I - I 4 C ] rsocoproate

1'

100

80 W +

? -

e

5

-"

enzyme during the assay. Preincubation of the tissue extract with all the componentsof the reaction mixture except for the substrate for 30 min at 37°C increased the initial rate of I4CO, formation %fold upon addition of the substrate, asjudged by comparing the 5-min time points on both curves in Fig. 3. After preincubation for 30 min, the activityremained constant for at least 15 min. Since itis known that activationof pyruvate dehydrogenase in adipose tissue extracts dependsupon the additionof MgCI,, we tested the possibility that activation of branched chain a-

60

0

4t t

+

c

p

40

a,

a

20

0 200pM

600pM

a-Ketolsovalerote

200pM

600.uM

DL-a-Keto-a-methyl-n-valerate

FIG. 2. Inhibition of the oxidation of [1-"Clpyruvate or aket~[l-'~C]isocaproate by unlabeled branchedchaina-keto acids. Adipose tissue extracts were preincubated in the presence of 5 mM MgCL for 30 min and assayed in the presenceof the indicated concentrations of unlabeled branched chain tr-keto acids in a reaction mixture containing either600 p~ [ I-''C]pyruvate or 200 p~ a-keto[l"Clisocaproate. In the absence of unlabeled branched chain u-keto 124 milliunits/g of acids the activities were: pyruvate dehydrogenase. tissue, and branched chaina-keto acid dehvdrogenase.7.56 milliunits/ g of tissue.

about the source of the "Cor since only the carboxyl carbon was labeled. However, "CO: may be derived from a-keto[U"C]-P-methyl-n-valerate by the initial oxidative decarboxylationand also by degradation of the putative product, amethylbutyryl-CoA. (The U-"C-labeled substrate was prepared from ~-[U-"C]isoleucine because L-[ l-'4C]isoleucine was not commerciallyavailable.)It was assumed for the purpose of calculation that '"CO:! would be derived from only the carboxyl carbon. The rate calculated, 3.57 milliunits/g, which is less than the observed rate of oxidation of a-keto[ 1I4 Clisocaproate or@-keto[ 1-"Clisovalerate,6.23 and 11.1 milliunits/g, respectively, is an upper limit for the oxidation of aketo-p-methyl-n-valerate.The low value of the rate is probably due to the low concentration of u-keto-/3-methyl-n-valerate, 25 ~ L as M compared to200 PM for the other branched chain tu-keto acids. Additional evidence suggesting the branched chain a-keto acids are oxidized by a single enzyme other than pyruvate dehydrogenase is presented in Fig. 2. Unlabeled branched chain u-keto acids greatly inhibited theoxidation of a-keto[l"Clisocaproate but not that of [l-14C]pyruvate (Fig. 2). Although the nature of the inhibition due to unlabeled cy-keto acids is not demonstrated, the results are consistentwith the idea that unlabeled branched chain a-keto acids inhibit aketo[ 1-"Clisocaproate oxidation becausethey areoxidized by the same enzyme, but they do not inhibit [1-'%]pyruvate oxidation because they are not oxidized by pyruvate dehydrogenase. When fresh extracts of adipose tissue were assayedfor their ability to oxidize a-ketoisocaproate, the rate of l4C02 formation was not 1inearl.v dependent on the length of the assay. Instead, the rate increased during assays longer than 5 min (Fig. 3). The increase can be explained by activation of the

Time, mln

FIG.3. Effect of preincubation on the activity of branched chain a-keto acid dehydrogenase. Adipose tissue extracts were either kept on ice or preincubated at 37'C for 30 rnin. The formation (130 p ~ during ) the indicated of "C02 from a-ket~[l-'~C]isocaproate intervalwasdeterminedasdescribeduner"Experimental I'rocedures." - 007

600/'

4

500-

400-

300-

1

0 0

2

4

Mogneslum chlortde

6

8

10

concentro11on. mM

FIG. 4. Dependence of pyruvate dehydrogenase and branched chain a-keto acid dehydrogenase activation on magnesium. Adipose tissue extracts containing the indicated concentrations of MgCI2 were incubated a t 37°C for 20 min and assayed either for p.yruvate dehydrogenase inthe presence of 6 0 0 [l-'4C]pyruvate, ~ ~ or for branched chain a-keto acid dehydrogenase in the presence of 200 PM a-ket~[l-'~C]isocaproate. Extracts incubated without added 16.3 MgClz hadthe following activities:pyruvatedehydrogenase, milliunits/g of tissue, and branched chain a-keto acid dehydrogenase, 0.746 milliunit/g of tissue.

Branched Chain a-Keto Acid

10

20

30

40

Dehydrogenase in Adipose Tissue

50

60

6189

7.4, and 1 mM dithioerythritol. Pretreatment of the tissues with insulin for 60 min doubled the activity of the enzyme in freshly prepared homogenates (Fig. 6). The effect of insulin persisted for 30 min when homogenates were stored at 0°C. Addition of insulin to homogenates of adipose tissue did not alter the activityof the enzyme. The effect of insulin on branched chain a-keto acid dehydrogenase activity was significantonly when assays were conducted a t low substrate concentrations (Table I). Insulin increased the activity of branched chain a-ketoacid dehydrogenase significantly when homogenates were assayed in the presence of 26 or 42 PM a-ketoisocaproate but not when the substrate concentration was as high as 69 or 79 PM. In each case enzyme activitiesin extracts of control and insulin treated tissues from the same animalswere compared. Enzyme activity was assayed 3 min after homogenizing each tissue. Disappearance of the effect of insulin at high substrate

Tune. rnin

FIG. 5. Effect of ATP on branched chain a-keto acid dehydrogenase activity. Adipose tissue extractswere combined withthe other components of the reaction mixture except for the substrate, preincubated at 37°C forthe indicatedtimes, andassayedfor branched chain a-keto acid dehydrogenase in the presence of 170 pM tu-keto[l-"C]isocaproate. ATP (50 nmol),creatinephosphate (5 pmol), and the creatine kinase (0.1 mg, 15 units) were added after preincubation for 30 min.

\

08-

063

keto acid dehydrogenase mightalso depend upon MgCI.. Adipose tissue was homogenized in 10 mM potassium phosphate buffer, pH 7.4, containing1 mM dithioerythritol and centrifuged to obtaina fat-poor infranatewhich was incubated at 37°C in the presence of 0, 1,5, or 10 m~ MgCl, for 20 min. Aliquots of the tissue extract were then transferred to the complete reaction mixture for assay of enzymatic activity in the ensuing 5 min (Fig.4). Preincubation with MgCln increased the activity of both the branched chain a-ketoacid dehydrogenase and pyruvate dehydrogenase, and the degree of activation was a function of the MgC1, concentration in the preincubation buffer. The time course of the activation was similar for both enzymes and the time course for branched chain a-keto acid dehydrogenase is shown in Fig. 5. Varying the concentration of calcium during the preincubation had little effect; addition of either 1 mM CaC12 or 2 mM EGTA along with 0 or 5 mM MgC12 did not substantially alter the activation of either enzyme (results not shown). In contrast, ATPdecreased the activity of branched chain a-keto acid dehydrogenase. In the experiment shown in Fig. 5, activation of the enzyme was essentially complete by 30 minwhen 0.1 mM ATP, along withanATPregenerating system consisting of 10 mM creatine phosphate and 15 units of creatine kinase was added. Branched chain a-keto acid dehydrogenase activity decreased by a factor of 4 within 11 min. Similar results have been reported for pyruvate dehydrogenase (25, 27). In another experiment, 0.1 mM ATP added without a regenerating system reduced branched chain a-keto acid dehydrogenase activity6-fold within 1min. No significant reduction in activity was producedby several othernucleoside GTP, triphosphates; a$-methyleneadenosine-5-triphosphate, or UTP, which were added a t concentrations of 0.1 or 1.0 mM. In order to study effect the of insulin on theenzyme, adipose tissue segments were incubated in Krebs-Ringer bicarbonate buffer containing 1 milliunit/ml of insulin and then homogenized in ice-cold buffer. Preliminary experiments indicated that the activity of branched chain a-ketoacid dehydrogenase was relatively stable for 30 min in homogenates prepared in ice-cold buffer containing 10 mM potassium phosphate, pH

Control

1

rq

Tlme, mln

FIG. 6. Effect of insulin on the activity of branched chain aketo acid dehydrogenase in adipose tissuehomogenates. Paired adipose tissue segments were incubated for 60 min in the presence or absence of insulin (1 milliunit/ml). Homogenates were prepared from the pooled tissues and stored on ice for the indicatedtimes.One minute prior to the assay aliquotsof homogenate were transferred to a shaking water bath (37°C).Assays were initiated by addition of 200 PI of the solution containing a-ket~[l-'~C]isocaproate (final concentration 21 PM) and other componentsof the reaction mixture.

TABLE I Effect of insulin on branched chain a-keto acid dehydrogenase activity in adipose fissue extracts Paired epididymalfat pads (150 to 300 mg) were incubated in Krebs-Ringer bicarbonate buffer with or without insulin (1 milliunit/ ml) for 20 min. The tissues were homogenized in 1.5 ml of buffer. The assay for branched chain a-keto acid dehydrogenase was initiated 3 min later by addition of 300 pl of homogenate (filtered through glass wool) to 200 pl of a solution containing the "C-labeled substrate and other components of the reaction mixture. The results are mean 2 S.E.; eight observations. concentration

Branched chain a-keto acid dehydrogenase activity

~ " _ _ _ _ ~

Control

Insulin "~~

PM

26 42 69 79

*

-~

I' - __

mrllrunrls/g trssue

0.36 0.04 0.55 2 0.07 1.64 f 0.15 1.46 0.22 -~

*

~~

0.54 k 0.09 0.86 2 0.08 2.21 f 0.19 1.59 k 0.11

4.05