Adenylosuccinate synthetase from Azotobacter vinelandii: Purification, properties and steady-state kinetics

ARCHIVES

OF

BIOCHEMISTRY

AND

Adenylosuccinate Purification,

BIOPHYSICS

184, 24-35 (1977)

Synthetase from Azotobacter vinelandii: Properties and Steady-State Kinetics’

GEORGE D. MARKHAM2 Department

of Biochemistry

and Biophysics,

AND

University Pennsylvania

GEORGE H. REED3

of Pennsylvania 19104

School of Medicine,

Philadelphia,

Received March 15, 1977 Adenylosuccinate synthetase has been purified to homogeneity from Azotobacter The purification method involves affnity chromatography on blue dextranSepharose, and hydrophobic chromatography, in addition to heat treatment, ammonium sulfate fractionation, and ion-exchange chromatography. The purified enzyme displays a single protein band after electrophoresis in the presence or absence of sodium dodecyl sulfate (SDS). Molecular weights of 110,000 and 54,000 are estimated by gel filtration and SDS gel electrophoresis, respectively. Steady-state kinetic measurements of the forward and reverse reactions and of the reaction in which arsenate replaces phosphate reveal a sequential mechanism with a fully random order of substrate addition in all cases. The maximal velocities of the reverse reaction and arsenolysis are virtually identical, and are approximately 10% of the maximal velocity for the forward reaction. In common with this enzyme from other sources, hadacidin is a potent competitive inhibitor with respect to aspartate (K, = 0.3 PM). Specific anions, e.g. nitrate and thiocyanate, are competitive inhibitors with respect to GTP; their effectiveness follows the Hofmeister series. Anion inhibition is synergized by GDP, but binding is exclusive with respect to guanylylimidodiphosphate, suggesting binding of the anions at the site normally occupied by the transferable phosphoryl group of GTP. vinelandii.

Adenylosuccinate synthetase [IMP:L-aspartate ligase (GDP), EC 6.3.4.41 catalyzes the reaction:

and the availability of substrates make this enzyme an attractive candidate for detailed studies of the mechanism of ligase action. However, the enzyme has proven M2+, GTP + IMP + L-aspartate GDP difficult to isolate in pure form from bac+ phosphate + adenylosuccinate terial sources (3-5). A homogeneous prepwhich is the penultimate step in the de aration of sAMP4 synthetase from rabbit muscle has been reported (61, but the ennouo synthesis of AMP. This reaction occupies a central position in nucleotide bio- zyme has only marginal stability. The synthesis and in cycling of purine nucleo- present paper describes a convenient purification scheme for sAMP synthetase from tides (1). The enzyme has a wide distribuAzotobacter vinelandii which gives reation, being found in microorganisms as sonable yields of stable, homogeneous enwell as in plants and a variety of animal zyme. The full steady-state kinetic mechtissues. A substantially elevated level of the enzyme has been found in neoplastic 4 Abbreviations used: sAMP, adenylosuccinate; tissues (2). Hepes, N-2-hydroxyethylpiperzine-N’-2-ethanesulThe reversible nature of the reaction fonic acid; Mes, 2-(N-morpholino)-ethanesulfonic 1 This work was supported by United States Public Health Service grant AM-17517. * Supported by United States Public Health Service predoctoral training grant 5TOl-GM374. 3 Address correspondence to this author.

acid, PMSF, phenylmethyl sulfonyl fluoride; GppNp, guanylylimidodiphosphate; DTE, dithioerythritol; SDS, sodium dodecyl sulfate; P-enolpyruvate, phosphoenolpyruvate; DEAE, diethylaminoethyl. 24

Copyright 0 1977by Academic Press, Inc. All rights of reproduction in any form reserved.

ISSN 9903-9961

A. uinelandii

ADENYLOSUCCINATE

anism is reported, including a comparison of the kinetic parameters of the reverse reaction with those for the reaction in which arsenate replaces phosphate, reactions whose details have heretofore remained obscure. The physical and kinetic properties of sAMP synthetase from Azotobacter vinelandii are compared to those of the enzyme from other sources. MATERIALS

AND

METHODS

Hepes, Mes, GTP, GDP, IDP, IMP, L-aspartic acid, PMSF, dithiothreitol, dithioerythritol, P-enolpyruvate (K+ salt), 2-mercaptosuccinate, and unspecified nucleotides were purchased from Sigma. sAMP was synthesized by the method of Hampton (7). Guanylylimidodiphosphate (GppNp) and 6-mercaptopurine riboside 5’-monophosphate were purchased from P&L Biochemicals. AMP used during the preparation was obtained from Fisher. Sephadex G-150 and DEAE-Sephadex A-50 were purchased from Pharmacia Fine Chemicals, Ltd. Blue dextran was coupled to Sepharose as described by Ryan and Vestling (8). DEAE-52 cellulose was purchased from Whatman Biochemicals, Ltd. Butyl-Sepharose was synthesized as described by Shaltiel (9). Hadacidin (N-formyl hydroxyaminoacetate) was a gift from Mr. Walter Gall, Merck & Co. Azotobacter uinelandii were grown as described by Schramm (lo), and stored frozen in 0.2 M phosphate buffer, pH 7.5, until use. Ammonium sulfate was purchased from Mallinckrodt or Schwarz/Mann. All other reagents were of the highest quality available and were used without further purilication. During the preparation, protein concentrations were estimated by assuming that E:% = 5.0 when AMP was present, and that E:$ = 10 in other cases. For the pure enzyme, an extinction coefficient, e:$, of 9.8 in 50 mM Hepes/KOH, pH 7.5, was determined by drying aliquots of enzyme and buffer to constant weight at 110°C. Kinetic studies and routine assays were performed on the 0.1 absorbance scale of a Zeiss PM1 spectrophotometer at 280 nm where there is a h of 11.4 x lo3 M-’ cm-l on the conversion of IMP to sAMP (11). Solutions were buffered with 50 mM Hepes/KOH, pH 7.0 containing 1 mM DTE. One unit of activity is defined as that amount of enzyme giving a change of one absorbance unit per minute at 280 nm in a l.O-ml reaction mix at 22°C when the substrate concentrations are: 5.0 mM L-aspartate, 0.2 mM GTP, 0.2 mM IMP, and 2 mM Mg(acetate),. One-centimeter path-length cells were used for all experiments. Velocities are expressed as change in absorbance at 280 nm per minute. Velocities referred to are initial velocities, except in cases involving anion inhibition where the steady-state ve-

SYNTHETASE

25

locity is reported (uide +a). In all cases, data points represent the average value of at least two experiments. Except where noted, reactions were initiated by addition of enzyme. RESULTS

Enzyme Purification

During the first three steps of the purification, through the DEAE-Sephadex chromatography, adenylosuccinate synthetase copurifies with AMP nucleosidase (EC 3.2.2.4). In order to obtain both enzymes from a single batch of bacteria, the nucleosidase activity was stabilized by addition of 2 mM AMP to all solutions during these steps (12). All steps in the preparation are performed at 0-4°C unless otherwise noted. Through the DEAE-Sephadex chromatography, all buffers contain 2 mM AMP, 0.1 mM EDTA, 0.5 mM dithiothreitol, and 3 PM PMSF. In subsequent steps buffers contain only 1 mM dithiothreitol. One and one-half kilograms ofdzotobacter vinelandii, stored frozen in 0.2 M phosphate buffer, pH 7.5 (total volume 2.4 liters), is thawed and the solution made 2 mM AMP, 0.1 mM EDTA, 0.5 mM dithiothreitol and 3 PM PMSF. The solution is blended for 45 s at low speed in a commercial blender and then the cells are broken by a single pass through a French press at 15,000-20,000 psi. The solution is placed in a stainless steel beaker and brought to 60°C by immersion in a water bath initially at 7O”C, allowing the bath to cool to 60°C as the solution heats. The preparation is maintained at 60°C for 10 min, and then cooled in an ice bath. Denatured protein is removed from the very viscous solution by centrifugation at 6000g overnight. The supernatant is diluted to a volume of 2.2 liters with 0.2 M potassium phosphate buffer, and 21.7 g of solid ammonium sulfate are added per 100 ml of solution. The solution is stirred for 15 min, and a small amount of precipitate is removed by centrifugation for 1 h at 104g. Solid ammonium sulfate is then slowly added (5.6g/lOO ml of supernatant). The solution is stirred for 15 min, and a large pellet is obtained by centrifugation at 104g for 2 h. This supernatant is discarded and

26

MARKHAM AND REED

the precipitate, which contains both the sAMP synthetase and AMP nucleosidase activities, is redissolved in 0.5 liters of 0.1 M Tris-HCl, pH 8.0, containing 2 mM AMP, 0.5 mM dithiothreitol, 0.1 mM EDTA, and 3 PM PMSF. The protein solution is then passed through an Amicon HlPlO hollow fiber dialysis apparatus against 4 vol of the Tris-HCl buffer containing 0.10 M NaCl. The dark brown protein solution is loaded onto a column of DEAE-Sephadex A-50 (5 x 40 cm) which has been previously equilibrated with the Tris buffer. The column is washed with 2 liters of the buffer containing 0.16 M NaCl followed by application of a linear gradient of 0.16 to 0.32 M NaCl in a total volume of 4.0 liters of the Tris buffer. sAMP synthetase is eluted at the end of the wash while AMP nucleosidase elutes near the end of the gradient. The fractions containing sAMP synthetase activity are pooled (total volume - 1 liter). The enzyme is concentrated to a minimal volume (100-200 ml) by diafiltration and dialyzed into 25 mM Tris-Cl, pH 7.5, buffer containing 1.0 mM DTE. Since the AMP nucleosidase has separated from the sAMP synthetase, AMP is not needed in the rest of this procedure. At room temperature, the protein is then passed through a column of blue dextran-Sepharose (2.5 x 20 cm) and the gel washed with the buffer until the absorbance at 280 nm of the effluent drops to that of the buffer. A linear gradient of IMP (0 to 7 mM in 0.5 liter) is then applied to elute the sAMP synthetase, and residual protein is removed with a 2 M salt wash. This chromatography is performed at room temperature since a sharper activity peak is obtained at 20 than at 4°C. Fractions containing the synthetase activity are pooled and loaded directly onto a column of butyl-Sepharose (1.8 x 10 cm> at 4°C. The column is washed with 25 mM potassium phosphate buffer, pH 7.0, until the effluent is free of nucleotide and the sAMP synthetase is eluted with a linear gradient of 0 to 0.4 M NaCl in 0.5 liter (Fig. 1A). Tubes containing the activity peak (specific activity 30) are pooled, and the enzyme is concentrated on a collodion bag

apparatus to at least 5 mg/ml. The activity is stable for several months when stored at 4°C under these conditions if the dithiothreitol concentration is raised to 5 mM. The progress of a typical purification is summarized in Table I. Since it is difficult to accurately assay the enzyme before the ammonium sulfate fractionation, percentage yields are calculated from the activity present after that step. This procedure routinely yields 20-30 mg of homogeneous enzyme from 1.5 kg of Azotobacter uinelandii.

Preparations with a specific activity of 30 are homogeneous as judged by the presence of a single protein band after polyacrylamide gel electrophoresis in the system of Davis (13) (Fig. 1B) (7.5 or 10% separating gel), by electrophoresis at pH 8.0 in 50 mM Tris-acetate buffer (7 or 10% gel), and by SDS gel electrophoresis in the system of Weber and Osborn (14). Identification Synthetase

of the Enzyme

as sAMP

The enzymic reaction may be monitored by uv difference spectroscopy, and the spectrum obtained of a reaction mixture minus a mix lacking only Mg(acetate), gives a ratio of absorbances at 282.5:280:270:262.5nm of 1.00:1.15:1.51:1.20 as predicted (1.00:1.14:1.50:1.14) for the difference spectrum of sAMP-IMP (15). Omission of any one component of the reaction mix results in no detectable reaction (~0.5% the rate of the complete system). Using a coupled assay system with pyruvate kinase and lactate dehydrogenase to measure the formation of GDP, no detectable cleavage of GTP occurs if either IMP or L-aspartate is omitted from the solution (< 2% the rate of the complete system). Furthermore, sAMP is degraded only when GDP, Mg*+, and phosphate (or arsenate) are all included in the reaction mix. Molecular

Weight Determination

The molecular weight of sAMP synthetase has been determined by gel filtration on Sephadex G-200 (1 x 61-cm column) in 0.1 M phosphate buffer, pH 7.0, containing 0.15 M KCl, 1 mM DTE with a Bow rate of

A. uinelandii

ADENYLOSUCCINATE

SYNTHETASE

27

A

ACTIVITY

(u/ml 1 30

6x3

15

1

0

profile of sAMP synthetase from the butyl-Sepharose FIG. 1. Part A shows the elution column. Solution were buffered with 25 mM potassium phosphate, pH 7.0, 1 mM DTE. At the arrow a gradient of 0 to 0.4 M NaCl in 0.5 liters of buffer was applied. sAMP synthetase is devoted by (-x-j and the absorbance at 280 nm by (-0-l. Part B shows a gel of the purified enzyme run in the systkm of Davis (13) with 10% separating gel and stained with Coomassie brilliant blue G.

ml/h. Lactate dehydrogenase (M, 130,000), creatine kinase (M, 82,600), and ovalbumin (M, 43,000) were included as standards. Enzymes were detected by their activities, and ovalbumin by absorbance at 280 nm. A plot of elution volume vs log molecular weight was linear for the standards and indicated a molecular weight of 110,000 (5 5 x 103) for sAMP synthetase. A molecular weight of 54,000 (2 2 x 103) was obtained by SDS gel electrophoresis (14) in the presence of 2mercaptoethanol, using chymotrypsinogen (M, 256001, ovalbumin (M, 43,000), and catalase (M, 60,000) as markers. 4.5

pH Optimum

and Substrate

Specificity

The basic reaction mixture for the following experiments contained 5.0 mM L-

aspartate, 0.20 mM IMP, 0.16 mM GTP, and 2.5 mM Mg(acetate),. In a 50 mM Mes/ 50 mM Hepes buffer system, with the pH adjusted by addition of KOH, the forward reaction shows a broad pH optimum from pH 6.5 to 7.4, with half maximal activities at pH 6.0 and 8.0. All further experiments were performed in 50 mM Hepes/KOH buffer, pH 7.0. No detectable reaction occurs in the absence of a divalent metal ion; at 1 mM metal ion, the rates of the forward reaction obtained with Mg(acetate), and Mmacetate), are equal, whereas Ca(acetate), gives approximately 20% of this rate. When GTP is replaced by 0.2 mM ATP, UTP, or CTP, no reaction is observed in a system in which 1% of the rate of the complete system could have been measured. IMP could not be replaced

MARKHAM AND REED

28

TABLE I PURIFICATION OF ADENYLOSUCCINATE SYNTHETASE FROM Azotobacter

Step Extract (1.5 kg cells) Heat step (NH&SO, fractionation DEAE-Sephadex pH 8.0 Blue dextran-Senharose Butyl-Sepharose

Volume (ml) 3.3 x 103 2.6 x IO3 700 835 93 61

Protein

(mn/ml) 65 22 30 2 N.D. 0.3

Total activity N.D.6 N.D. 1600 1000 732 603

Specific activity 0.05 0.6 30

uinelandii

Yield (%) =lOO 63 46 38

n Purification is estimated as total protein reduction before the salt fractionation, activity increase in subsequent steps. * N.D., Not determined.

by hypoxanthine, inosine, UMP, GMP, 6mercaptopurine-riboside 5’-phosphate or 6-chloropurine-riboside 5’-phosphate at 0.2 mM concentration. At a concentration of 5.0 mM, L-glutamate, L-glutamine, p-alanine, and m-2-mercaptosuccinate were ineffective in substituting for L-aspartate. The substrate specificity for the reverse reaction was examined using a basic system containing 25 mM phosphate, 100 PM GDP, 21 PM sAMP, 2 mM Mg(acetate), in 50 mM Hepes/KOH, pH 7.0, 1 mM DTE. No detectable reaction (i.e., < 1% the rate of the complete system) occurred in the absence of any one component of the mixture. It was found that neither 0.2 mM IDP nor 0.15 mM 2-aminopurine-riboside 5’-diphosphate could replace GDP. Fluorophosphate (25 mM) and phosphite (30 mM) were inactive as substitutes for phosphate, although substantial activity was observed in the presence of 10 mM arsenate (uide infru). Steady-State

Kinetics

The steady-state kinetic mechanism for uinelandii sAMP synthetase was elucidated both by conventional inhibition methods and by the method of Rudolph and Fromm (16). In the latter procedure, the changing fixed substrates are varied in a constant ratio, and the slopes and intercepts of the double-reciprocal plots are replotted against the reciprocal of the changing fixed substrate concentrations. The shapes of the replots reveal the kinetic mechanism. Figure 2 shows the Lineweaver-Burk olots for each of the three substrates of

Azotobacter

Puritication” 3.8 10 124 6210 and by specific

the forward reaction when levels of the other two substrates are varied at a constant ratio. In all cases the lines at different levels of the changing substrates clearly intersect, thus demonstrating a sequential mechanism. Replots (not shown) of the slopes and intercepts of these lines versus the reciprocal of the concentration of the changing fixed substrates are all nonlinear, and none intersect the origin, indicating a fully random order of substrate addition (16). A fully random, sequential mechanism predicts that an inhibitor which is competitive with respect to one substrate will be a noncompetitive inhibitor with respect to the other two substrates (171, and this is indeed observed. GDP and GppNp are competitive inhibitors with respect to GTP (K, = 30 and 15 PM, respectively) and noncompetitive with respect to both IMP and aspartate. sAMP and 6-mercaptopurine riboside 5’-phosphate are both compejitive with respect to IMP (K, = 2.5 and 22 PM, respectively) and noncompetitive with respect to both GTP and aspartate. Furthermore, hadacidin and P-enolpyruvate are competitive inhibitors with respect to aspartate (Ki = 0.3 PM and 0.2 mw respectively) and noncompetitive with respect to both nucleotide substrates. The kinetic data are consistent with a fully random, sequential mechanism for the forward reaction. The product inhibition patterns indicate that the reverse reaction must either be fully random or ordered in such a fashion that the first and third substrates add in a random fashion and the second in compulsory order (19).

A. uinelandii

ADENYLOSUCCINATE

Kinetics

FIG. 2. Double-reciprocal plots for the forward reaction. In all cases solutions contained 50 mM HepeslKOH, pH 7.0, 1 mM DTE, and 1.0 mM Mgcacetate),, T = 20°C. Part A shows results obtained with GTP as variable substrate. The concentrations of the changing fixed substrates are: (-x -) 0.12 mM IMP, 0.98 mM L-aspartate. (-A-) 23.2 FM IMP, 196 PM L-aspartate. (-0-1 11.6 FM IMP, 98 pM L-aspartate. (-0-j 6.9 pM IMP, 58 PM L-aspartate. Part B shows IMP as the variable substrate. The concentrations of other substrates are: (-x-) 154 PM GTP, (0.98) mM L-aspartate; (-0-) 50.9 PM GTP, 0.33 mM L-aspartate; C-A-) 30.8 PM GTP, 196 pM L-aspartate; (-0-j 15.4 pM GTP, 98 PM L-aspartate. Part C shows L-aspartate as variable substrate. The other substrate concentrations are: (-x-) 77 /LM GTP, 71 PM IMP, (-0-j 38 /LM GTP, 35 ,uM IMP; C-A-, 19 +M GTP, 17.6 /.&M IMP; (-0-j 12 /.LM GTP, 11 /.LM IMP: (-0-j 7.7 PM GTP, 7.1 PM IMP.

SYNTHETASE

29

of the Reverse Reaction

Although the forward reaction has been examined for the enzyme from other sources (3, 4, 16, 181, no information regarding kinetic parameters of the reverse reaction or arsenolysis is available. Therefore, the kinetics of the reverse reaction and of the arsenolysis reaction were examined using the method of Rudolph and Fromm (16). Several of the LineweaverBurk plots for these data are shown in Figs. 3 and 4 (for the phosphorolysis and arsenolysis, respectively). For arsenolysis, replots of the slopes and intercepts are all nonlinear and do not intersect the origin, indicating a fully random order of substrate addition. For the phosphorolysis reaction, both the slope and intercept replots of the l/V versus l/[GDPl data appear to be linear. A linear slope suggests a partially ordered mechanism in which GDP adds second, while sAMP and phosphate may bind either first or third. However, it is not possible to saturate the enzyme with respect to phosphate due to substrate inhibition, and this factor may influence the replots. In view of the arsenolysis reaction which shows fully random substrate addition, and of equilibrium binding measurements which show that MnGDP and free GDP both bind to the enzyme in the absence of other substrates (201, a fully random pattern of substrate addition is the only scheme consistent with all the results. Although substrate inhibition is observed by both phosphate and arsenate at high concentrations (> 30 mM), the extrapolated maximal velocities for the two reactions, phosphorolysis and arsenolysis, are virtually identical. Since substrate addition in both directions is fully random, it is possible to obtain the affinities of each substrate for free enzyme (K,) from the common point of intersection of the double-reciprocal plots at various levels of the changing fixed substrates (19). In a rapid equilibrium mechanism, the affinity of the substrate for the enzyme liganded with the other two substrates (Km) is obtained from the extrapolated slope/intercept ratio at infinite concentrations of the other sub-

MARKHAM

AND

REED

strates (16). These constants are summarized in Table II. Inhibition

100

200 ,M-’

FIG. 3. Double-reciprocal plots for the reverse reaction. All solutions contained 50 mM Hepesl KOH, pH 7.0 buffer, 1 mM DTE, 2.0 mM Mg(acetate)*. Part A shows phosphate (potassium salt) as the variable substrate. The concentrations of the changing fixed substrates are: (-x-j 200 pM GDP, 42 /.LM sAMP; (-0-j 60 /AM GDP, 12.6 /AM sAMP; (-A-) 40 PM GDP, 8.4 PM sAMP. Part B shows results when GDP is the variable substrate. The other substrate concentrations are: (-x-j 20.0 m&r phosphate, 21 PM sAMP; (-0-j 12.0 mM phosphate, 12.6 PM sAMP; (-A-) 8.0 mM phosphate, 8.4 pM sAMP. Part C shows results obtained with adenylosuccinate as the variable substrate. The concentrations of other substrates are: C-X-) 20.0 mM phosphate, 200 PM GDP; (-0-) 10.0 m&r phosphate, 100 PM GDP; (-A-) 5.0 mM phosphate, 50 PM GDP.

Experiments

. In addition to the results described above, the enzyme is inhibited in a competitive manner with respect to aspartate by succinate (Ki = 1.8 mM), 2-mercaptosuccinate (Ki = 3.2 mM), oxalacetate (Ki = 0.5 mM), and phosphate (Ki = 5 mM). While fumarate at 5 mM causes slight noncompetitive inhibition with respect to aspartate, maleate is a competitive inhibitor (K, = 0.9 mM), suggesting that the carboxyl groups of aspartate are oriented in a predominantly cis conformation when bound to the enzyme. At 5 mM concentrations, the following compounds give no detectable inhibition (with L-aspartate at 0.1 mM>: L-alanine, p-alanine, L-glutamine, and L-glutamate. AMP is a competitive inhibitor with respect to IMP (Ki = 100 PM). At 10 PM IMP, no detectable inhibition is caused by 2 mM n-ribose 5-phosphate or 0.2 mM 6chloropurine riboside 5’-monophosphate. Specific anions are inhibitors of sAMP synthetase from other sources, inhibition being competitive with respect to GTP (21, 22). This pattern of inhibition is also observed with the Azotobacter vinelandii enzyme. Figure 5A shows a Dixon plot for several of these anions. The Dixon plot exhibits an initial region in which l/V is linear with anion concentration, followed by a region in which the plot curves strongly. The curvature suggests multiple modes of interaction. At high concentrations, the overall order of effectiveness of the anions tested is C1,CC02- > SCN- > I- > NO,- > SO,p2 > NOz- > Cl- > HC02- with acetate giving no detectable inhibition. This order appears to follow the Hofmeister series for lyotropic anions, as was noted for rabbit muscle sAMP synthetase- (22). At low anion concentrations (< 5 mM> and high substrate concentrations the order of effectiveness of the anions is altered, with NO,- becoming the best inhibitor. This further suggests that the anions interact with the protein at more than a single site.

A. uinelandii

ADENYLOSUCCINATE

FIG. 4. Double-reciprocal plots of the arsenolysis reaction. Solutions contained 50 mM Hepes/KOH, pH 7.0 buffer, 1 mM DTE, 2.0 mM Mg(acetate),. Part A shows arsenate (sodium salt) as the variable substrate. The concentrations of the other substrates are: (-x-j 22 FM sAMP, 147 PM GDP; (-0-j 8.8 /.LM sAMP, 59 &LM GDP: (-A-) 4.4 /AM sAMP, 29 pM GDP. Part B shows GDP as the variable substrate. Other substrate concentrations are: f-x-) 10.0 mM arsenate, 21 pM sAMP; (-0-l 6.0 mM arsenate, 13.2 pM sAMP; (-A-) 4.0 mM arsenate, 8.4 pM sAMP. Part C shows results obtained when

SYNTHETASE

31

Inhibition by NO,- is competitive with respect to GTP, and noncompetitive with respect to aspartate and IMP (Fig. 6). SCN- is also a competitive inhibitor with respect to GTP. When the concentrations of all substrates are decreased together, curvature in the Dixon plot becomes noticeable at lower concentrations of anion. When the concentration of each substrate is varied individually, it becomes apparent that the degree of curvature of the Dixon plot is unaffected by the aspartate concentration at constant concentrations of GTP, and IMP. On the other hand, decreasing the concentration of either nucleotide substrate enhances the curvature of the Dixon plot. Similarly, Lineweaver-Burk plots for IMP and GTP curve strongly upward as the anion concentration is increased, suggesting positive cooperativity of nucleotide binding in the presence of anions. Doublereciprocal plots for aspartate remain linear in the presence of anion. A similar effect is seen with threonine deaminase from Bacillus subtilis, where threonine binding is not cooperative in the absence of inhibitor, but addition of the competitive inhibitor isoleucine results in apparent positive cooperativity in both substrate and inhibitor binding (23). A further similarity to the threonine deaminase system is the appearance of a transient in the progress curve of the reaction in the presence of inhibitor (24). When the reaction is initiated by addition of sAMP synthetase, a lag in obtaining the steady-state velocity is observed, and the duration of the lag increases linearly with anion concentration (Fig. 5B). Within the range 0.1 to 10 times their K,,, values, the duration of the lag phase is independent of substrate concentrations. Incubation of enzyme with anion and two substrates followed by initiation of the reaction with the third substrate reveals that the lag phase is abolished if the enzyme adenylosuccinate is the variable substrate. The other substrate concentrations are: (-x-) 10.0 rnM arsenate, 147 PM GDP; (-0-j 5.0 mM arsenate, 74 +M GDP; C-A-1 3.0 mM arsenate, 59 FM GDP; (-0-J 1.5 mM arsenate, 29 WM GDP.

32

MARKHAM

AND

TABLE KINETIC

n-----ard

CONSTANTS

reaction

K,,

&

&I

II

FOR ADENYLOSUCCINATE

SYNTHETASE”

Reverse reaction (phosphorolysis)

15 pLM 48 /.LM 13 /AM 30 /.bM 0.5 rnM 90 /.LM 32 AA,,, unitslminlmg 2.8 ~mol/min/mg

GTP IMP L-Aspartate V

REED

sAMP GDP Phosphate

Arsenolysis

KU

KS

-l/.&M

~/AM 20 ELM

48 /.LM 15 rnM 8mM 3.2 AA,,, units/min/mg 0.28 ~mol/min/mg

sAMP GDP Arsenate

KItI

-1 /.LM -5 ~.LM 53 /.LM 34 PM 5rnM 2.5 rnM 3.2 AAzs,, units/min/mg 0.28 Fmol/min/mg

a All solutions contained 50 mM Hepes/KOH pH 7.0, 1 mM DTE. For the forward reaction solutions contained 1.0 mM Mg(acetate),. In the reverse reaction and arsenolysis solutions contained 2.0 mM Mg(acetate)*. T, 20°C. KD is the dissociation constant of the substrate from the free enzyme. K, is the Michaelis constant for the substrate (see text).

1 A

6

HCOj I 0

I

1

I

I

3o

(ANION)

I mM

I GO

I

I

I 90

1

180

7 sec. GO

(ANION1

5. Part

mM

A shows a Dixon plot of anion inhibition of sAMP synthetase. All solutions contained 0.11 mM GTP, 0.17 rnrd IMP, 2.3 mM L-aspartate, 1 mM Mg(acetate)* in 50 mM Hepes/KOH, pH 7.0, 1 mM DTE. The symbols represent (-0-j potassium trichloroacetate, (-0-l KSCN, (-A-) KNO,, (-0-j sodium sulfite, (-x-j sodium formate. Part B shows the lag phase duration, 7, plotted against anion concentration. 7 is defined as shown in the inset. The experimental conditions are those described in part A. The symbols represent: C-X-) potassium trichloracetate; (-0-j KSCN, (-0-j KNO,, (-0-l sodium sulfite. FIG.

A. uinelandii

*

2

G

ADENYLOSUCCINATE

10

FIG. 6. Double-reciprocal plots of sAMP synthetase inhibition by nitrate. Solutions contained 50 mM Hepes/KOH, pH 7.0, 1 mM DTE, 1 mM Mgcacetate),. Part A shows competition of NO,with GTP. The concentrations of fixed substrates are 70 PM IMP, 0.5 rnru L-aspartate. The symbols represent: f-x-) no inhibitor; (-0-j 7.8 mM KNO,; (-A-) 11.7 mM KNO,; (-0-j 15.2 mM KNO,. Part B shows inhibition with respect to IMP. The fixed substrate concentrations are 70 pM GTP, 0.5 mM Laspartate. The symbols represent C-X-) no inhibitor; (-0-j 7.6 mM KNO,; (-A-) 11.7 mM KNO,; (-0-j 15.2 mM KNO,. Part C shows inhibition with respect to L-aspartate. The concentrations of the other substrates are 67 PM GTP, 68 pM IMP. The symbols represent f-x-) no addition; (-0-j 7.6 mM KNO,; (-A-) 15.2 mM KNO,.

SYNTHETASE

33

has been preincubated with IMP. Moreover, no lag in obtaining the new steady state is observed if the anion is added to the reaction in progress. In order to determine if the inhibitory anions altered the aggregation state of sAMP synthetase, the molecular weight of the enzyme was determined by gel filtration in 50 mM Hepes/KOH alone and in the additional presence of 0.1 M potassium acetate (which does not inhibit) or 0.1 M KSCN (which is a strong inhibitor). Identical molecular weights were found in all cases, indicating that the anion effects are not due to alteration of the aggregation state of the enzyme. The coincidental appearance of a transient and of apparent cooperativity is reminiscent of the predictions of the slow transition model (251, which shows that an interconversion of enzyme forms of,differing activity that is slower than turnover, combined with nonrapid equilibrium substrate binding, can lead both to transients and to apparent cooperativity. The lag phase results suggest that there is a slow interconversion between an inactive form of the enzyme, which the anions stabilize, and an active form that is stabilized by IMP. When two compounds of very different structure are competitive inhibitors with respect to the same substrate, it is important to ascertain if they are both bound to the enzyme simultaneously. Figure 7 shows a Dixon plot of NO,- inhibition at various levels of GDP. Lines at various concentrations of GDP intersect above the abscissa, demonstrating that the inhibition by NO,- is potentiated by the presence of GDP (6). From the point of intersection, an interaction coefficient of 0.17 is obtained, corresponding to an approximately sixfold synergism in the binding of the two inhibitors. On the other hand, in the presence of GppNp, which is also competitive with respect to GTP, parallel lines are observed in the Dixon plot of NO,inhibition, demonstrating that the binding of NO,- and GppNp is mutually exclusive (26). It appears likely that NO,- binds at the site normally occupied by the transferable phosphoryl group of GTP.

34

MARKHAM

AND

REED

even at pH 6.0 and moves quite rapidly in the above electrophoretic system. The electrophoretic properties of the Azotobacter vinelandii sAMP synthetase resemble those of the enzyme from Escherichia coli. The pH optimum for the forward direction of the Azotobacter vinelandii enzyme is midway between those of E. coli and rabbit muscle enzymes (6, 27). In the forward direction, a fully random, sequential (ND;1 KM steady-state kinetic mechanism has also been reported for the E. coli and human FIG. 7. Yonetani-Theorell plot showing effects placental enzymes, among others (3, 4, of GDP and guanylylimidodiphosphate on inhibition 16, 18). The two bacterial enzymes have by nitrate. All solutions contained 70 PM GTP, 51 similar K, values for their substrates in pM IMP, 0.5 mM L-aspartate, 1 mM Mg(acetate)* in the forward direction, both binding IMP 50 rnr.4 Hepes/KOH, pH 7.0,l mM DTE. The symbols show C-X-) no addition; (-A-) 30 PM GDP, (-0-j 59 approximately an order of magnitude pM GDP, (-0-j 22 PM GppNp. The steady-state more tightly than the rabbit muscle envelocity is plotted and reactions were initiated by zyme. It is noteworthy that the affinities addition of enzyme. of both IMP and MgGTP are enhanced when other substrates are liganded to the DISCUSSION enzyme (cf. Table II). This suggests that substrate-induced conformational changes The purification procedure detailed here yields stable, homogeneous enzyme of spe- in the enzyme may be important in achievcific activity comparable to that reported ing the proper disposition of the system for the crystalline rabbit muscle enzyme. for catalysis. Although the sAMP synthetase reaction The critical step in this simple purification has long been known to be reversible (281, of a homogeneous bacterial sAMP syntheno kinetic study of the reverse reaction tase is the blue dextran-Sepharose affinity has been reported previously. Substrate chromatography. Electrophoretic analysis addition is fully random for both the phosof enzyme eluted from blue dextran-sephphorolysis and arsenolysis reactions, and arose with 10 mM IMP pulses gave bandthe extrapolated maximum velocities for ing patterns similar to those obtained with identical, KC1 washes, and that eluted with a 10 the two reactions are virtually being approximately an order of magnimM GTP pulse was only slightly improved. Furthermore, the enzyme was not dis- tude slower than that for the forward In both cases, the affinity of placed from the resin by 10 mM ATP or by direction. sAMP for the enzyme is significantly 10 mM NADH. Elution from the affinity greater for free enzyme, while GDP is resin with an IMP gradient yields enzyme bound more tightly to the fully liganded which is 80 to 90% pure. The impurities enzyme. Again, the data suggest the imremaining after the blue dextran-sephaportance of conformational changes during rose step are readily removed by the butylthe reaction which are similar, although Sepharose chromatography. apparently not identical, for phosphoroThe sAMP synthetases from rabbit muslysis and arsenolysis. cle and from Azotobacter vinelandii have In common with sAMP synthetase from similar molecular weights and subunit other sources (29, 30), the Azotobacter vistructures. However, the isoelectric points nelandii enzyme is inhibited by the asparof the two enzymes differ substantially. tate analog, hadacidin, with a Ki several The rabbit muscle enzyme does not bind lower than K,n for to anion-exchange resins, even at pH 8, orders of magnitude aspartate, suggesting substantial homoland moves very slowly during electrophoogy in the amino acid binding site. Inhibiresis in the system of Davis (13). On the other hand, the Azotobacter vinelandii en- tion of sAMP synthetase by NOB- is zyme is tightly bound to DEAE-cellulose strongly enhanced by GDP, while NO,-

1 d

59vM

GDP

A. uinelandii

ADENYLOSUCCINATE

and GppNp appear to compete with each other. These findings are consistent with the proposal that the planar anion NOsoccupies the binding site of the transferrable phosphoryl group, mimicking the structure of a metastable intermediate in forming a transition state analog complex enzyme-MgGDP-IMP-L-Asp-NO,(21). It should be noted that at least for the more lipophilic anions another mode of inhibition may also be operative, particularly at low substrate and high anion concentrations. Inhibition of an sAMP synthetase by phosphoenolpyruvate has not previously been noted. Since significant inhibition occurs in the tens of micromolar concentration range, this is a potential physiological control mechanism. However, it is likely that the activity of sAMP synthetase is regulated more by the steady-state levels of various nucleotides than by a highly specific control mechanism. In view of the relatively slow maximal velocity, perhaps this should not be too surprising, even though the enzyme does act at an important anabolic branch point. sAMP synthetase from Azotobacter vinelandii satisfies several of the requirements for extensive physical study. The activity is stable indefinitely when the enzyme is stored at 4°C in medium containing thiol-protecting reagents such as dithiothreitol. The enzyme is very soluble, as concentrations in excess of 100 mg/ml are readily achieved. The relatively slow turnover rate of the enzyme may facilitate investigations of pre-steady state kinetics. Finally, the yield of homogeneous enzyme from the present scheme is sufficient to support experiments which require large quantities of enzyme. ACKNOWLEDGMENT The authors thank Dr. Vern L. Schramm of Temple University for his generous supply of Azotobacter uinelandii, and for his aid during development of this enzyme purification. REFERENCES 1. BUCHANAN, J. M., AND HARTMANN, S. C. (1959) Aduan. Enzymol. 21, 199-261. 2. JACKSON, R. C., MORRIS, H. P., AND WEBER, G. (1975) Biochem. Biophys. Res. Commun. 66,

SYNTHETASE

35

526-532. 3. ISHII, K., AND SHIIO, I. (1970) J. Biochem. (Tokyo) 68, 171-1’76. 4. NAGY, M., DJEMBO-TATY, M., AND HESLOT, H. (1973) Biochim. Biophys. Acta 309, l-10. 5. STAYTON, M., AND FROMM, H. J. (1976) Abstr. 172nd Amer. Chem. Sot. Meeting, 155. 6. MUIRHEAD, K. M., AND BISHOP, S. H. (1974) J. Biol. Chem. 249, 459-464. 7. HAMPTON, A. (1962)J. Biol. Chem. 237,529-535. 8. RYAN, L. D., AND VESTLING, C. S. (1974) Arch. Biochem. Biophys. 160, 279-284. 9. SHALTIEL, S. (1974) in Methods in Enzymology, Vol. 34 (Jakoby, W. B., ed.), pp. 126-140. Academic Press, New York. 10. SCHRAMM, V. L. (1974) Anal. Biochem. 57, 377382. 11. ATKINSON, M. B., MORTON, R. K., AND MURRAY, A. W. (1964) Biochem. J. 92, 398-404. 12. SCHRAMM, V. L., AND HOCHSTEIN, L. I. (1972) Biochemistry 11, 2777-2763. 13. DAVIS, B. J. (1964) Anal. N.Y. Acad. Sci. 121, 404-427. 14. WEBER, K., AND OSBORN, M. (1969) J. Biol. Chem. 224, 4406-4412. 15. TORNHEIM, K., AND LOWENSTEIN, J. M. (1972) J. Biol. Chem. 247, 162-169. 16. RUDOLPH, F. B., AND FROMM, H. J. (1969) J. Biol. Chem. 244, 3832-3839. 17. FROMM, H. J. (1966) Biochim. Biophys. Acta 139, 221-230. 18. VAN DER WEYDEN, M. B., AND KELLEY, W. N. (1974) J. Biol. Chem. 249, 7282-7289. 19. FREIDEN, C. (1957) J. Amer. Chem. Sot. 79, 1894-1896. 20. MARKHAM, G. D. (1977) Ph.D. Thesis, University of Pennsylvania, Philadelphia. 21. MARKHAM, G. D., AND REED, G. H. (1975) FEBS Lett. 54, 266-268. 22. BISHOP, S. H., FISCHER, H. E., GIBBS, K. L., AND STOUFFER, J. E. (1975) Fed. Proc. 34, 548. 23. HATFIELD, G. W., AND UMBARGER, H. E. (1970) J. Biol. Chem. 245, 1742-1747. 24. HATFIELD, G. W., RAY, W. J., JR., AND UMBARGER, H. E. (1970) J. Biol. Chem. 245, 17481754. 25. AINSLIE, G. R., SHILL, J. P., AND NEET, K. E. (1972) J. Biol. Chem. 247, 7088-7096. 26. YONETANI, T., AND THEORELL, H. (1964) Arch. Biochem. Biophys. 106, 243-251. 27. LIEBERMANN, I. (1956) J. Biol. Chem. 223, 327339. 28. MILLER, R. W., AND BUCHANAN, J. M. (1962) J. Biol. Chem. 237, 485-490. 29. SHIGEURA, H. T., AND GORDON, C. N. (1962) J. Biol. Chem. 237, 1937-1940. 30. RUDOLPH, F. B., AND CLARK, S. W. (1975) Fed. Proc. 34, 508.