Rat liver adenosine triphosphate: adenosine monophosphate phosphotransferase activity. I. Purification and physical and kinetic characterization of adenylate kinase 3.

Abstract Rat liver adenylate kinase III (ATP:AMP-phosphotransferase, EC 2.7.4.3) has been purified and characterized. The enzyme was purified 1500-fold to a final specific activity of 1000 µmoles of ADP produced per min per mg of protein at 25°. The preparation was homogeneous by polyacrylamide gel disc electrophoresis and at 10 mg per ml was essentially a single component by analytical ultracentrifugation. Molecular weight studies with Sephadex G-150 column chromatography revealed a Vo:Ve ratio of 1.94 which corresponded to a molecular weight of 46,000 for the enzyme. A range of Svedberg values as a function of protein concentration was obtained by ultracentrifugation and established the presence of two molecular weight forms of the enzyme; extrapolation to zero protein concentration indicated two S020, w values of 1.23 and 3.52. With the extrapolated value of the diffusion coefficient of 4.8 x 10-7 cm2 sec-1 and the partial specific volume of 0.74 calculated from the amino acid analysis, molecular weights of 23,000 and 68,000, respectively, were obtained. Frictional and axial ratios were found to be 1.1 and 4.0, respectively. We propose that rat liver adenylate kinase III is a globular protein existing as monomer, dimer, or trimer in very rapid equilibrium. Amino acid analysis revealed a total of 216 amino acids with a calculated minimum molecular weight of 23,400. The protein was found to be high in glutamic acid, alanine, aspartic acid, and leucine residues; it was low in histidine, methionine, half-cystine, and phenylalanine residues. Initial velocity studies revealed a narrow specificity for adenine nucleotides. The respective Km values for ATP, dATP, and dGTP were 0.12, 0.77, and 1.77 mm with respective Vmax values of 16,000, 40,000, and 25,000 moles of triphosphate min-1 mole-1 of enzyme. The only monophosphate acceptor, with a Km of 0.12 mm was 5'-AMP. The Km for ADP was 0.18 mm with a Vmax of 10,700 moles of diphosphate min-1 mole-1 of enzyme. It would appear that rat liver adenylate kinase III slightly favors the forward reaction (conversion of ATP).

* This work was supported by Grants AM-08350, CA-10439, and C&10906-08 from the National Institutes of Health, and Grant P-202 from the American Cancer Society.
Nucleotide-transphosphorylating enzymes occur within most forms of life. Many of these enzymes are not specific for the triphosphate moiety, or not specific for the monophosphate moiety or not for both (l-6).
Adenylate kinase activity also may appear in several electrophoretically distinct forms (17)(18)(19)(20)(21). Three bands were observed with agarose-acrylamide gel electrophoresis of several rat tissues (19) and four bands were found with isoelectrofocusing of rat liver (21). The four bands separated by isoelectrofocusing were called I, II, III, and IV according to their isoelectric points of pH values 5.9, 7.0, 7.6, and 8.2. In the latter study (21) we reported the loss in hepatomas of the predominant liver isozyme form, adenylate kinase III.
This isozyme was also the form which was responsive to diet and hormones.
In this communication we report the puriiication and physical and kinetic characterization of rat liver adenylate kinase III. For this equation a partial specific volume of 0.74 was used which was calculated from the amino acid analysis, according to the method of Cohn and Edsall (26).
A frictional ratio, faverage /fminimum, was calculated according to Tanford (27). From the value of the frictional ratio an axial ratio was also determined (27).
Animals-CFN male rats were purchased from Carworth Farms, Rockland County, New York. The animals were maintained on lab chow until they were between 200 to 300 g.
Enzymatic Assays-Adenylate kinase activity was measured in both directions utilizing coupled enzyme systems. In the forward direction ADP formation was coupled with phosphoenolpyruvate, pyruvate kinase, and lactate dehydrogenase; NADH oxidation was continuously monitored at 340 rnp. In the reverse direction, ATP formation was coupled with glucose, hexokinase, and glucose g-phosphate dehydrogenase; NADPf reduction was continuously measured at 340 rnp. Details of these kinetic assays have been previously published (21).

Puri)ication
of Adenylate Kinase III Preparation of Cytosol-Rats were fasted 36 hours before killing.
The animals were decapitated, exsanguinated, and the livers removed and placed in cold 12.5 InM sucrose-l mu cysteine. All of the following procedures were performed at O-5". The liver tissue was cooled, blotted, weighed, and homogenized with 2 volumes of cold 12.5 mM sucrose-l mM cysteine in a Waring blendor.
The whole homogenate was centrifuged in a Spinco model L ultracentrifuge at 30,000 rpm for 2 hours. The supernatant was used as the cytosol.
pH Fractionation-The cytosol was subjected to pH frac- tionation by a modification of the procedure of Noda and Kuby (28). To lower the pH of the cytosol from 6.8 to 3.0, 3 N HCl was added very rapidly with stirring. The solution was stirred for 10 min. Added very slowly (over a 5-min interval) with stirring was 1 N NaOH to restore the pH to 6.8. The solution was stirred for 20 min and then centrifuged at 16,300 x g for 20 min. The precipitate was discarded. The supernatant fraction was concentrated by ultrafiltration with a UM-10 membrane.
Sephadex G-75 Chromatography-The concentrated pH 3 supernatant was layered on a column, 5.6 X 45 cm, containing Sephadex G-75 equilibrated with 5 mM sodium phosphate buffer-1 mM fl-mercaptoethanol at pH 7.2. The same buffer was used to elute the enzyme from the column.
Fractions (8 ml each) were collected and assayed for enzymatic activity.
Fractions containing 90% of the adenylate kinase activity were combined and concentrated by ultrafiltration with a UM-10 membrane.
IsoelectrofocusingLThe concentrated G-75 fraction was applied to an electrofocusing column containing carrier ampholytes of pH 5 to 8. P-Mercaptoethanol or cysteine (1 mM) was added to the column media. The electrofocusing procedure has been described previously (21). At the end of Le electrofocusing run, 3-ml fractions were collected am. assayed for enzymatic activity. Fractions containing 90% of adenylate kinase III activity were combined and concentrated by ultrafiltration. A UM-10 membrane was used.
Rechromatography on Sephadex G-75-The concentrated electrofocused fraction was dialyzed with 30 volumes of 5 mM sodium phosphate buffer at pH 7.2 in the ultrafiltration cell with a PM-10 membrane. The washed concentrate was layered on a Sephadex G-75 column, 3 x 77 cm, ecluilibrated in 5 mM sodium phosphate buffer at pH 7.2. The enzyme was eluted wit,h the same buffer and 3-ml fractions were collected and assayed for enzymatic activity. Fractions containing 90% of t,he adenylate kinase activity were combined and concentrated by ultrafiltration with a UM-10 membrane. A summary of the purificat.ion scheme is given in Table I. The final specific activity was 700 pmoles of ADP produced per mm per mg of protein at 25". The over-all recovery was 30% with a 1600-fold increase in purification. The last Sephadex step gave little increase in purification, but was necessary t,o remove ampholytes remaining from electrofocusing and ultrafiltration dialysis.

Enzyme Homogeneity
Polyacrylamide Disc Gel Electrophoresis-Electrophoresis of the purified adenylate kinase III revealed one protein band (Fig. 1) which contained all the adenylate kinase activity on the gel.
Analytical Ultracentrijugation-One peak was observed in the analytical ultracentrifuge when adenylate kinase III was centrifuged at 10 mg per ml (Fig. 2). However, skewing of the slower sedimenting portion of the peak is indicative of a lower molecular weight component as will be discussed below.
Adenylate kinase III appeared to be homogeneous by electrophoresis and essentially one component by ultracentrifugation at lower protein concentrations.
Physical Properties of Adenylate Kina.se III Sephadex G-160 Analysis of Molecular Weight-The molecular weight of adenylate kinase activity from rat liver was studied with liver cytosol extracts from normal, 4%hour fasted, 4%hour fasted-16-hour glucose refed, diabetic, and diabetic-insulintreated rats All of these procedures caused large changes in viva   (21,29). Partially purified preparations and the final homogeneous preparation were examined in the presence and absence of ATP, ADP, and Mg++. In all studies, the major (>90%) component has a V,:V, ratio of 1.94 which corresponded to a molecular weight of 46,000 on a calibrated column (Fig. 3). Infrequently, small but significant amounts of adenylate kinase activity peaks were also observed which corresponded to molecular weights of approximately 21,000 and 160,000. The latter may be caused by nonspecific aggregation. Analysis by Ultracentrijugation-A range of Svedberg values was obtained by ultracentrifugation of solutions of 3 to 30 mg per ml of purified adenylate kinase III (Fig. 4). The Svedberg values increased rapidly in the range from 3 to 10 mg per ml and decreased slowly from 10 to 30 mg per ml. Extrapolation to zero protein concentration revealed two &,, values of 1.23 and 3.52, FIG. 6. Sedimentation velocity of adenylate kinase III at 30 mg of protein per ml. Direction of sedimentation is from left to right. Photograph was made 112 min after attaining a speed of 60,096 rpm at 20" with a double sector cell and the AN-D rotor. The solvent was 5 mM phosphate, pH 7.2; reference solution was water in this experiment. The main heavier peak is 2.3 S and the lighter shoulder is 0.9 S. a condition which is virtually diagnostic of interacting syst,ems involving association-dissociation equilibria with very rapid forward and backward reactions (30). Diffusion coefficients were determined from 10 to 30 mg per ml of purified adenylate kinase (Fig. 5). The slope of the plot of D,"O,, versus protein concentration was zero. With the D&,, of 4.8 x 10-r cm2 set-1, the molecular weights of 23,000 and 68,000 were calculated for the two Svedberg values (Table II). At a high protein concentration of 30 mg per ml, both the lower as well as the higher forms were visible (Fig. 6). This is taken .to confirm the reversible nature of the equilibrium between monomer and polymer. The partial specific volume was calculated from the amino acid analysis to be 0.74 (Table II). The frictional and axial ratios were found to be 1.1 and 4.0, respectively (Table II). It would appear that rat liver adenylate kinase III is a globular protein which exists as a monomer, dimer, or trimer (possibly higher polymers) which are in very rapid equilibrium.
Amino Acid Analysis-Two separate preparations of purified rat liver adenylate kinase III were used for the amino acid analysis (Table III). Each preparation of 4 mg per ml was hydrolyzed for 48 and 72 hours with 1 volume of 6 N HCl at 110' in sealed glass tubes which had been made anaerobic with Nz. Such incubations at 110" for 24 hours yielded incomplete hydrolysis. Calculated from the hydrolysate were 45 dicarboxylic acid and 29 diamino acid (including histidine) residues. This difference in charged groups may contribute to the iso-   (21). A total of 216 amino acid residues were determined which gave a calculated minimum molecular weight of 23,400 (Table III). This was in good agreement with the lower molecular weight of 23,000 which was obtained from analytical ultracentrifugation.
Kinetic Properties of Adenylate Kinase III Substrate Speciifcity-Initial velocity studies on purified adenylate kinase III revealed a rather narrow specificity for adenine nucleotides (Table IV). The respective K, values for ATP, dATP, and dGTP were 0.12, 0.77, and 1.77 mM with V max values of 16,000, 40,000, and 25,000 moles of triphosphate min-l mole+ of enzyme.
The enzyme did not use GTP, ITP, UTP, or CTP as triphosphate donors. Adenylate kinase III used only 5'-AMP as monophosphate acceptor with a K, of 0.12 mM. Cyclic 3', 5'-AMP, dAMP, 2'-AMP, 3'-AMP, CMP, GMP, dGMP, IMP, TMP, or UMP did not serve as monophosphates. The K, for ADP was 0.18 mM with a 17,, of 10,700 moles of diphosphate min+ mole-l of enzyme. Other diphosphates were not examined for specificity because of the limitations of the reverse assay system.

DISCUSSION
Rat liver adenylate kinase III has been purified 1500-fold to a final specific activity of 1000 pmoles of ADP produced per min per mg of protein at 25'. The preparation was observed to be homogeneous by polyacrylamide gel disc electrophoresis and was essentially a single component by analytical ultracentrifugation.
Several unique differences are apparent when the physical protein data of the adenylate kinase from bovine liver mitochondria, rabbit muscle, and Bakers' yeast are compared with the rat liver adenylate kinase III (6)(7)(8)(9)(10)(11).
Diffusion constants are near 10 x lo-+ cm2 see-l for the bovine liver mitochondrial and rabbit muscle enzymes, 7 x 10-V cm2 set-l for Bakers' yeast enzyme, and 4.8 x lo-' cm2 set-' for rat liver III enzyme. With Svedberg constants of 2.49, 2.30, 2.96, and both 1.2 and 3.5 for bovine liver mitochondrial, rabbit muscle, Bakers' yeast, and rat liver III enzymes, respectively, the molecular weights were calculated to be 21,500 and 21,000 for the bovine liver mitochondrial and rabbit muscle adenylate kinases, 40,000 for the yeast enzyme, and at the three different weights of 23,000,46,000, and 68,000 for the rat liver adenylate kinase III.
It would appear that all of the purified adenylate kinases have similar minimum molecular weights, but the yeast and rat liver III forms are capable of aggregating at higher protein concentrations.
Polymers of the bovine liver mitochondrial enzyme may have been missed since it was examined at protein concentrations (0.4 to 4.0 mg per ml) which were below those required to observe the aggregated forms of the rat liver III enzyme (above 8 mg per ml).
The bovine liver mitochondrial, rabbit muscle, Bakers' yeast, and rat liver III enzymes all used ATP + AMP and ATP + dAMP.
The bovine liver mitochondrial adenylate kinase also reacted with,ITP + AMP; the yeast enzyme catalyzed reactions between dATP + AMP, GTP + AMP, and ATP + GMP; the rat liver III enzyme also reacted with dGTP + AMP.
The forward versus reverse reaction was almost equal with the bovine liver mitochondrial enzyme and the rabbit muscle enzyme. Maximal velocity studies with the yeast enzyme and rat liver III enzyme showed the forward reaction (conversion of ATP) to be favored.
Large differences were observed in the apparent Km values as K, values for all the adenine nucleotides with the bovine liver mitochondrial adenylate kinase were above 1 mu; the K,,, values for the rabbit muscle and rat liver III enzymes were from 0.12 to 0.50 mM; K,,, values for yeast enzyme were from 0.052 to 0.27 mM.
When considered together, the physical data from analytical ultracentrifugation and molecular sieve chromatography indicate a reversible aggregation system for rat liver adenylate kinase III. O-OO-6b