Mar 4, 2018 - following modifications: ACC was precipitated at 5.5% PEG instead of. 5%, and the protein was precipitated at 40% saturation of ammonium.
Vol. 269, No. 9, Issue of March 4, pp. 6859-6865, 1994 Printed in U S A .
THEJOURNAL OF BIOIOGICAL CHEMISTRY 0 1994 by The American Society for Biochemistry and Molecular Biology, Inc
Isolation and Characterizationof a Novel Acetyl-coA Carboxylase Kinase from Rat Liver* (Received forpublication, September 20, 1993, and in revised form, November 10, 1993)
k Habib Mohamed, Wei-Yong Huang, Wanzhi Huang,K. V. Venkatachalam, and Salih J. Wakils From the Verna and Marrs McLean Department of Biochemistry, Baylor College of Medicine, Houston, Texas77030
stimulated by CoA (5).Hardie andco-workers (6-10) also have Acetyl-coAcarboxylase is regulatedallosterically employed many different proteinkinases to phosphorylate by citrate and covalently by phosphorylation/ a ACC. The most notable of these are an AMP-activated protein dephosphorylation mechanism. We have isolated and purified from rat livers a novel kinase that phosphokinase from rat livers and another protein kinase, acetyl-coA rylates and inactivates the carboxylase. This kinase is carboxylase kinase 2,from rat mammaryglands (10). The bound to the carboxylase and can be eluted in salt-richAMP-activated kinase is reported to have a molecular weightof solution. The nativekinaseexists as highmolecular 63,000(8)and is not specific for ACC since it phosphorylates weight aggregates of a subunit that has a molecular the p-hydroxy-p-methylglutaryl-CoAreductase, the hormoneweight of 40,000. The phosphorylation sites of the car- sensitive lipase/cholesterol esterase, and possibly otherenboxylase were determined after tryptic and cyanogen ACC zymes (9, 11).Both kinases phosphorylate and inactivate bromide digestions of 32P-labeled carboxylase and sepaand are CAMP-independent enzymes (10). However, phosphoration of the peptidesby various chromatographic pro- rylation of the carboxylase by acetyl-coA carboxylase kinase 2 cedures. Amino acid analyses of the phosphopeptides of the carboxylase in waysthat changes the kinetic properties showed that the S e P and Serlam residues were sites the are similar to those caused by phosphorylation mediated by the of phosphorylation. -eating the phosphorylated carCAMP-dependent kinase (10). The AMP-activated protein kiboxylase with the Mn2+-dependent acetyl-coA carboxyl(site 3), SerlZo0(site 11, and possibly ase phosphatase 2 removed the phosphate and reacti- nasephosphorylates Ser1215 ( l l ) , whereas CAMP-dependent proteinkinaseand vated the carboxylase. These results suggest that both AMP-independent acetyl-coAcarboxylase kinase 2 both phosthis kinase and the acetyl-coA carboxylase phosphatase 2) and Ser’200 (4,10,12). Protein kinase C 2 act at the same site(s) in the acetyl-coA carboxylase phorylate Ser77 (site can also phosphorylate the Ser77of the ACC molecule (4, 10, molecule. Citrate dramatically inhibits the kinase-medi12). These threesites (Ser77,Ser79, and SerlZo0) apparently are atedphosphorylationofthecarboxylase,suggesting that the allosteric modification and activation by citrate linked to theactivity modulation of ACC (4, 9, 12). render the phosphorylationsites inaccessible to the kiIn our studies of the regulation of ACC by phosphorylation, nase and therefore maintain high carboxylase activity. we noted that affinity-purified ACC undergoes “autophosphoThis observation indicates that there is a close interplay rylation” independent of added kinase. Further investigation on and phosphorylation of therevealed that this “autophosphorylation” varied fromone between the citrate effect carboxylase in regulatingits activity. preparation to another, suggesting the presence of a n ACCbound kinase. Herein, we report the isolation and characterization of a novel protein kinasefrom rat livers that phosphorylatesandinactivates ACC. The phosphorylation of ACC mediated by this kinase is dramatically inhibited by citrate. This kinase does not require CAMP,AMP, or CoA. We propose that this kinase, bound to its substrate ACC, is involved in the activity modulation of ACC both in vivo and in vitro.
Acetyl-coA carboxylase (ACC),’ a cytoplasmic enzyme, catalyzes the carboxylation of acetyl-coA to form malonyl-CoA, which is the rate-limiting step in the synthesis of long-chain fatty acids (1).Citrate, which is produced in the mitochondria and transported to the cytoplasm, is the precursor of acetylCoA and acts asa feed-forward allosteric activatorof ACC (2). EXPERIMENTALPROCEDURES The mode of citrate activation is not clear; however, its action retired-breeder rats werepurchased does lead topolymerization of the protein into aggregates with Materials-Spraguefemale molecular weights greaterthan lo7 (3). The carboxylase is also from Harlen Co. Polyethylene glycol(PEG)8000, phenylmethanesulfonyl fluoride, biotin, phosphorylase b, histone, benzamidine, leupeptin, regulated by covalent modification of theprotein by a aprotinin, trypsin inhibitor,CAMP protein kinase (from porcine heart), phosphorylatioddephosphorylation mechanism.SeveralkiCAMP proteinkinase peptide inhibitor,CAMP, and all otherchemicals nases have been found to phosphorylate the protein and reduce and biochemicals were purchased from Sigma. [14C1Bicarbonate and and its activity(4). Kim and co-workers (5)isolated from rat livers ammonium sulfate werepurchasedfromICNBiochemicals, ts2PlATP from DuPont NENor Amersham Corp. a AMP-independent kinase that phosphorylates and inactiPreparation ofAcetyl-CoA Carboxylase-ACC was prepared fromthe vates ACC. The phosphorylation mediated by this kinase is livers of fasted and refed rats as described previously (13) with the modifications:ACC was precipitated at 5.5% PEG instead of * This work was supported by Grant GM-19091 from the National following 5%, and the protein was precipitated at 40% saturation of ammonium Institutes of Health and a grant from The Clayton Foundation. The part by the payment sulfate. The unbound (avidin flow-through) material was collected becosts of publication of this article were defrayed in of page charges. This article must thereforebe hereby marked “adver- fore the column was washed and saved as the source of carboxylase tisement” in accordance with 18 U.S.C. Section 1734 solely to indicate kinase (seebelow). Acetyl-coA Carboxylase Assay-ACC was assayed by using a modithis fact. $ To whom correspondence should be addressed. Tel.: 713-798-4528; fication of the [‘*CIbicarbonate fixation assay asdescribed by Thampy and Wakil(14).One unit of activity is defined as 1m o l of malonyl-CoA Fax: 713-796-9438. The abbreviations used are: ACC, acetyl-coAcarboxylase;PEG, formedmin at 37 ”C. Specific activityis defined as 1 unitimg protein. polyethylene glycol; PAGE, polyacrylamide gel electrophoresis; HPLC, Carboxylase Inactivation Assay-The inactivation assaymixture (1 high performance liquid chromatography. ml) contained 50 nm HEPES(pH 7.9, 10 nm MgCI2, 1 nm dithio-
Acetyl-coA Carboxylase Kinasefrom Rat Liver
threitol, 1 mg of bovine serum albumin, 0.1 rn ATP, and 20 pg of affinity-purifiedACC. The reaction was started by adding 0.1-5 pg of kinase to the mixture. After incubating the mixture for 3 0 4 0 min at 37 "C, an aliquot containing 0.2 pg ofACCwas removed and assayed for carboxylase activity by using the procedure described above. of ACC in the CarboxylaseP h o s p ~ o ~ Assay-Phospho~lation ~~~on presence of [y-32PlATPand kinase was performed as described forthe inactivation assay, exceptthat bovine serum albumin was omitted and 250 pg of affinity-purifiedACC, 5 pg of kinase, and 0.1 rn of [32PlATP (1 pCi/nmol) were used. After incubation, the reaction was stopped by adding Laemmli solubilization buffer (15). The ACC was separated in 5% SDS-PAGE (13) and visualized by autoradiography. For quantitation, the Carboxylase band was excised from the dried gel, cut into pieces, soaked in a solution containing 0.5 ml of water and 4.5 ml of scintillation mixture, and the radioactivity was determined. Alternatively, after the ACC was incubated with the kinase, aliquots were withdrawn and layered onto Whatman 3-mm paper squares (2 cm x 2 cm) that had been presoaked in 500 p~ cold ATP and dried. The paper discs were dried at 40 "C and washed three times in cold 10%trichloroacetic acid(5-10 ddisdwash). The paper discs were then immersed in ethanol, rinsed with ether, and air-dried, and the radioacti~ty was counted in a liquid scintillation spectrometer (16). Other Methods-The amount of protein was estimated by using the dye-binding method (17), and silver staining of the gels was done as previously described (18). Phosphoserine Analysis and Phosphopeptide Sequencing-The 3pPlabeled ACC generated by the aforementionedprocedure was separated from the reaction mixture by precipitating it with ammonium sulfate. For phosphoserine analysis, the radioactive protein was hydrolyzed with 4 N HCl a t 100 "C for 2 h. An aliquot (10-15 pl)of the acid hydrolysate was spotted with standard phosphoamino acids (phosphoserine, phosphothreonine, and phosphotyrosine) onto a Whatman cellulose thin-layer (0.5 m m ) chromatographic plate and developed in a high voltage (500 V) electrophoretic system in two steps. In the first step, a pyridine:acetic acidH20 (1:990) solvent system (pH 3.5) was used for 3.5 h. In the second step, the thin-layer ch~matographicplate was developed with a solvent system (pH 2.1) offormic acidacetic acid water (2:890) for 4.5 h. The thin-layer plate was air-dried and sprayed with ninhydrin to identify the amino acids. The radiolabeled amino acids were visualized by autoradiography. For phosphopeptide sequencing, the radioactive protein (5 mg) was digested with trypsin and cyanogen bromide(19-22). For tryptic digestion, [32PlACC(1 mg/ml in 0.2 N N-ethylmorpholineacetate, pH 7.5) was digested with trypsin a t an enzyme to substrate ratio of 1:lOO at 37 "C for 24 h. Digestion was stopped by injecting the digest directly into an HPLC column for purification, For a-chymotrypticdigestion and cyanogen bromide cleavage, [32P]ACC(1 mdml in 0.2 N N-ethylmorpholine acetate, pH 7.6) was digested with a-chymotrypsin at an enzyme to substrate ratio of 1:200 at 37 "C for 5 min. One dropof acetic acidwas then added, and the digest was freeze-dried. The dried product was dissolved in 300 pl of 70% formic acid, 10 mg of cyanogen bromidewas added, and themixture was lefi to react at mom temperature for 24 h
concentrated, and transferred to an Immobilon-P membrane and sequenced as described above.
RESULTS ~ ~ ~ ~ co a~ t~ ~e ot yn~ ~ - a~ ~ rA Kinase-The ~ o carbox~ ~ ylase kinase was first detected as an ACC-associated protein after the ACC had been precipitated with5.5% PEG and fractionated with 40% ammonium sulfate (see "Experimental Procedures"). This association persisted even after ACC bound to the avidin-Sepharose column. However, whenthe bound ACCavidin column was washedwith avidin column buffer containing 0.5 M NaCl, most of the kinase wasreleased i n the eluate. Eluates from several preparations (equivalent to 60 rat livers) were pooled, and enough ammonium sulfate was added to achieve 50% saturation. The precipitated protein was separated by centrifugation, redissolved in kinase buffer (50 m~ Tris-HC1 (pH 7.Q 1 RIM EDTA, 1 RIM dithiothreitol, 0.1 m~ phenylmethanesulfonylfluoride, and 5% glycerol), and dialyzed against the same buffer for 12 h. The dialyzed solution was centrifugedto remove denatured proteins and loaded onto a DEAE-Bio-Gel column (I) (2.5x 25 cm) that was equilibrated with kinasebuffer. After extensive washing with kinase buffer, most of the kinase was elutedin 0.5 M NaCl inthe same buffer. Ammonium sulfate was added to the eluate to 50% saturation, and the precipitated protein was removed by c e n ~ ~ g a t i o n , redissolved in kinase buffer, and dialyzed against the same buffer for 12 h. The dialyzed solution was clarified by centrifugation, assayed for kinase activity, and loaded onto a second DEAE-Bio-Gel column (11) (2.5 x 25 cm). The proteins were eluted by a linear gradient consistingof 200 ml eachof kinase buffer and of 0.5 M NaCl in the same buffer, and fractions of 8.3 ml each were collectedand assayed forkinase activity by using the [32P]ATPassay procedure (see "Experimental Procedures"). As shown in Fig. 1 (A a n d B ) , the kinase eluted in fractions 5-30. However, we found that the early fractions (5-11)contained nearly pure kinase, whereas later fractions were contaminated with many proteins as shown by Coomassie Blue staining of gels (data not shown). Further purification of the kinase was counterproductive, since the purified enzyme was unstable and each step introduced resulted in rapid loss of kinase activity. Hence, we optedto use fractions 5-11 (Fig. 1A) in one-step purification by using a DEAE-Affi-Gel Blue column. These fractions were pooled, and the protein was precipitated with ammonium sulfate at 50% saturation. The precipitated in kinase protein was isolated by c e n t ~ ~ g a t i o redissolved n, (21). buffer, and dialyzed against the same buffer for 4 h. The diaThe tryptic peptides were separated by HPLC by using a reversea small DEAE-AEi-Gel Blue phase Vydac CIS (5 pm, 300 & column (0.4 x 25 cm) on a Waters lyzed mixture was loaded onto Gradient model 6000A chromatograph. The solvent system used was column (3 ml). The column was washed with 30 ml of kinase 0.1%trifluoroacetic acid (solventA) and 95% CH&N with 0.1%trifluo- buffer, and the proteins were eluted in successive washings roacetic acid (solvent B). Fractions (1.0 ml each) were collected, and with 5 ml eachof kinase buffer containing50 mM NaCl, 10 m~ their radioactivity was measured by using a liquid scintillation spec- ADP, and 2 M NaCl. Fractions were collected separately and trometer. The fractions containing the radioactive peptides were pooled, tested for kinase activity(Fig. 2,A and B ) . Analysis of the 50 concentrated into a small volume, and rechromatographed on the same kinase on column. ARer rechromatography, the fractions containing the radioac- m~ NaCl eluate of the DEAE-Mi-Gel-Blue-purified SDS-PAGE showed a single protein band with an estimated tive peptides were pooled and concentrated. The peptides wereconverted to phenylthiocarbamyl derivatives (23) and were purified by molecular weight of 40,000 (Fig. 3). However, when the active using the same HPLC column and solvent system. After purification, kinase fractions were concentrated and analyzed by chromathe fractions containing the radioactive peptides were pooled and con- tography on a Superose 6-FPLC inthe presence of 0.1 M "iscentrated to a volume of 0.1 mf. The contents were transferred to an HCI (pH 7.5)and 0.1 M NaCl, most of the kinase eluted in the Immobilon-P membrane, and the peptides were sequenced by using an AppliedBiosystemsmodel 470A gas-phase protein sequenator (19). void volume, suggesting that the kinase exists as high molecuPhenylthiohydantoin derivatives were identified on-line by using an lar weight aggregates under these conditions (data not shown). AppliedBiosystemsmodel 120A phenylthiohydantoin analyzer (20). Generally, the highly purified kinase was labileand lost activThe radioactive amino acids were identified according tothe method of ity within5 days when stored at -70 "C. However, the partially Wang et al. (21). purified fractions that eluted at about 150 m~ NaCl in the The radioactive peptides generated by the chymotryptic and cyano- DEAE Bio-Gel column (11) (fractions 12-25) were stable for gen bromide digestions were separated by HPLC by using a reversephase Vydac C,, (5 pm, 300 A) column (0.45 x 25 cm) and the solvent several months whenstored at -70 "C. As summarized in Table system used for the tryptic peptides (see above). Two radioactive pep- I, this procedure overall purifiedthe kinase by more than 500tides (C-CNl and C-CN2) wereobtained, which were separately pooled, fold.
Acetyl-coA Carboxylase Kinase from Rat Liver
FIG.1. A, DEAE-Bio-Gel (11) elution profile of acetyl-coA carboxylase kinase. Afterfractionationthroughthe DEAEBio-Gel (I) column (0.5 M NaCl eluate), the proteinswereloaded onto the second DEAE-Bio-Gel (11) column, washed with kinase buffer, and eluted with a linear gradient of 200 ml each of kinase buffer containing 0 and 0.5 M NaCI, respectively. Fractions (8.3 mleach)were collected, and aliquots (50 p1) were tested for ACC phosphorylation activity by using the filter paper method (see ’Experimental Procedures”). E , phosphorylation activity of the DEAE-Bio-Gel (11) fractions of acetylCoA carboxylase. Aliquots (50 pl each)of the earlier fractions of the NaC1-gradient eluates were testedfor ACC phosphorylation activity in the presence of [“2PlATP and ACC (25pg).Thereactionwas stopped after 60 min of incubation a t 37 “C by adding SDS-solubilization buffer. Thesampleswereanalyzedin SDS-PAGE (5%). The protein bands were visualized by stainingwithCoomassie Blue(notshown)and autoradiography. Lane 1, DEAE-Bio-Gel (11) column load; lane 2, column load without ACC; lane 3, ACC withoutcarboxylasekinase; lanes 4-10, ACC plus kinase from column fractions 5-11, respectively.
1 2 3 4 5 6 7 8 9 10
- 125 kDa
A. 1 2 3 4 5
B. 6 7 8 9 1 0
1 2 3 4 5 6 7 8 9 1 0
- 125 kDa
General Properties of the Acetyl-coA Carboxylase KinaseIncubating ACC with the purified kinase in the presence of [y-32PlATPresulted in time-dependent incorporation of radioactivity into the carboxylase (data not shown). Under these conditions, maximum incorporationof 32P into theACC protein
FIG.2. DEAE-Affi-Gel Blue elution fractions of the acetyl-coA carboxylase kinase. Theactive carboxylase kinase fractions5-11 (Fig. lA)were pooled, and the proteins were isolated and loaded onto aDEAE-Affi-GelBlue column, which was then eluted in succession with 5 ml each of kinase buffer containing 50 m~ NaCl, 10 mMADP, and 2 M NaCl (see text). Aliquots (50 pl each) were used to assay for ACC phosphorylation a s described in Fig. 1. A, the gel stained with Coomassie Blue; E , autoradiogram of the same gel. Lanes 1 4 , control without ACC; lanes 6-10, with ACC. Lanes 1 and 6, DEAEAffGGel Blue column load; lanes 2, 3, 7, and 8 , 5 0 m~ NaCl eluate; lanes 4 and 9, 10 mM ADP eluate; lanes 5 and 10, 2 M NaCl eluate.
(M,= 260,000) was achieved after 60 min of incubation. The maximum incorporation of 32Pwas quitevariable, being in the range of 0.1-0.5 mol of Pi/mol of carboxylase (data not shown). The K,,, values of the carboxylase kinase for ATP and for ACC were 30 p~ and 300 m, respectively. Another protein band (M,
Acetyl-CoA Carboxylase Kinase
from Liver Rat
centration (Fig. 7). Other anions, such as NaCI, did not profoundly affect ACC phosphorylation(Fig. 7). We could not attribute this citrate-mediated inhibition of ACC phosphorylation to the complexing of citrate with M$“, which would reduce the effectiveness of the Mg2+.ATPcomplex in thereaction. For the citrate-mediated inhibition occurred even when the Mg2‘ concentration was raised to 35 mM, which exceeds the citrate concentration. Moreover, citrate, at the concentration used in the ACC phosphorylation assay, had no effect on the phosphorylation of phosphorylase bkinase by CAMP-dependentprotein kinase when i t was used as a control. -40 kDa The data presented inFig. 6 also show that thepresence of CAMP (0.2 mM), AMP (0.5 mM), or CoA (0.4 m ~ in) the ACCphosphorylation reaction mixture did not affect the phosphorylation of ACC by this kinase. Previously, these factors were considered necessary for the activity of their respective kinases in phosphorylating ACC (5-8). Moreover, CAMP-dependent protein-kinase peptide inhibitor (16) did not inhibit nor did the presence of palmitoyl-CoA enhance ACC phosphorylation (Fig. 6). Hardie and co-workers (26) reported that palmitoyl-CoA enhanced the phosphorylating activity of purified AMP-activated kinase; however, they attributed this increaseto activa- .tion of a contaminating kinase-kinase activity in their prepaFIG.3. SDS-PAGEanalysis of DEAEXffi-Gel Blue-purifiedace- ration. No such activation was noted in our kinase preparation. tyl-coA carboxylase kinase. Kinase samples (100pl) were obtained In addition, the kinase described herein did not phosphorylate a s described in Fig. 2, analyzed in SDS-PAGE (lo%),and stained with silver nitrate(18).Lane 1, DEAE-”Gel Blue column load; lanes 2 and phosphorylase b or histone (data notshown). Altogether, these 3,50rn NaCl eluate; lanes 4 and 5 , 10 mM ADP eluate (see legend of observations and the size of the kinase subunit strongly sugFig. 2). gest that the kinase we isolated is not the one isolated by Lent and Kim (5) nor is it the AMP-activated kinase reported by = 125,000), which was also labeled with32Pby the kinase, was Hardie and co-workers (6, 7). identified as a proteolyzed fragment of ACC based on the posiIncubating the carboxylase with [32P]ATPin the absence of tive reaction of the bands with anti-ACC antibodies ina West- the carboxylase kinase resulted in the incorporation of a small ern blot assay (data not shown). but relatively significant amount of radioactivity into theproKinase-mediated phosphorylation of ACC resulted in loss of tein (Fig. 6, lane2 ), suggesting that this “autophosphorylation” carboxylase activity, which also was time-dependent (Fig. 4). is due to the bound kinase. This incorporation varied from The inactivation of ACC was ATP-dependent; after incubating preparation to preparation and depended upon the extent to 60 min with the kinase and ATP, ACClost morethan 75% of its which the carboxylase-bound avidin-Sepharose column was activity (Fig. 4). Incubating the ACC with the kinase in the washed with thesalt-rich buffer during the preparation ofACC absence of ATP resulted insome loss of activity (