Regulation of D-Fructose l-Phosphate Kinase by Potassium Ion*

Abstract Purified Aerobacter aerogenes d-fructose 1-phosphate kinase (ATP:d-fructose 1-phosphate 6-phosphotransferase), an enzyme which does not exhibit cooperative kinetics, was activated by K+, Rb+, and NH+4, inhibited by Li+, and not significantly affected by Na+ or Cs+. The enzyme exhibited some activity when assayed in tetramethylammonium-glycylclycine buffer even in the absence of other monovalent cations, suggesting that the requirement for monovalent cations is not absolute. K+ increased the reaction velocity about 4-fold in the presence of excess ATP and d-fructose-1-P, and about 10-fold when both substrates were present at concentrations near their Km values (0.3 mm). K+ also exerted a much greater activating effect under conditions of inhibiting ATP levels than when Mg++ was present at a sufficiently high level to prevent inhibition by ATP. The Ka for K+ (about 3 mm) was not affected by the Mg++ or ATP concentration, but was increased by decreased levels of d-fructose-1-P. K+ decreased the Km values for both ATP and d-fructose-1-P. These kinetics are consistent with Michaelis-Menten theory and suggest a simple model in which the enzyme exists in two active forms, one form predominating in the presence of K+ and the other predominating in its absence.

Because the intracellular concentration of K+ in microbial cells is dependent upon their metabolic state (4)(5)(6), and because K+ is required by many enzymes for maximal activity (7)(8)(9), the possibility exists that K+ may serve a regulatory role in * This investigation was supported by Research Grant AI 08066 from the National Institute of Allergy and Infectious Diseases, United States Public Health Service; Journal Article 4970 from the Michigan Agricultural Experiment St,ation. microbial metabolism. n-Fructose-g-P kinase from various sources is activated by both K+ (10)(11)(12)(13) and NH,+ (ll- 16), but n-fructose-l-P kinase from Bacteroides symbiosus apparently is not activated by monovalent cations (17). In this paper we describe the effect of K+ on n-fructose-l-P kinase from A. aerogenea PRL-R3.
K+ increases the V,,,, increases the affinity of the enzyme for ATP and n-fructose-l-P, and partially overcomes inhibition by excess ATP. A model is proposed for the activation of n-fructose-l-P kinase by K+ which is consistent with the observed kinetics. MATERIALS AND METHODS 2MaterialsThe n-fructose-l-P kinase used in these studies was the 315fold purified preparation (specific activity, 47.0) described previously (3). Bovine serum albumin was obtained from Sigma, and tetramethylammonium hydroxide (25% in water) from Mallinckrodt Chemical Works. Other chemicals were obtained as described previously (3).
Removal of Activating JTonovalent Cations from Reaction Componentsprior to assay, the n-fructose-l-P kinase was diluted 160-fold with a solut,ion of bovine serum albumin (2 mg per ml) in 0.1 M tetramethylammonium-glycylglycine buffer (pH 7.5). The residual (NH&SO4 and sodium acetate in the diluted enzyme solution amounted to approximately 8 PM and 0.1 PM, respectively, in the final reaction mixture. Bovine serum albumin, cr-glycerophosphate dehydrogenasetriose phosphate isomerase, and n-fructose-l, 6-dip aldolase were freed from activating monovalent cations by passage through a column of Sephadex G-25 equilibrated with 0.1 M tetramethylammonium-glycylglycine buffer (pH 7.5). Fractions were collected by eluting with the same buffer. Sodium ions were removed from ATP and NADH by dissolving the nucleotides (sodium salts) in 0.2 M tetramethylammonium-glycylglycine buffer (pH 7.5) and passing the solutions through a column of Sephadex G-15 equilibrated with the same buffer. The concentrations of the nucleotides in the eluted fractions were determined by absorbance at 259 nm. Barium n-fructose l-phosphate was treated with Dowex 50 (H+) to remove barium ions, and the resulting free acid was neutralized with (CHz).,NOH. n-Fructose l-phosphate concentrations were determined enzymically (3).
pH measurements during preparation of the buffer and neutralization of n-fructose l-phosphate were made with a Sargent miniature combination electrode (manufactured by Jena Glass Works, Mainz, Germany), which has an extremely low electrolyte leak rate. Possible contamination of the reagents with KC1 from the electrode was determined with a Coleman flame Issue of June 25, 1970 V. Sapico and R. L. Anderson  The velocity is expressed as nanomoles of Dfructose-1,6-di-P formed per min.
photometer. The K+ contamination from this source was found to be negligible ( <5 PM in the final reaction mixture). D-Fructose-l-P Kinuse Assay-The basic reaction mixture (0.15 ml) contained 0.5 pmole of ATP; 1.0 pmole of MgC&; 1.0 pmole of n-fructose-l-P; 0.05 pmole of NADH; 10 kmoles of tetramethylammonium-glycylglycine buffer (pH 7.5); nonlimiting amounts of D-fructose-l, 6-diP aldolase, triose phosphate isomersse, and oc-glycerophosphate dehydrogenase; and limiting amounts of D-fructose-l-P kinase. The amounts of the coupling enzymes necessary to assure that the reaction rates were linear with the n-fructose l-phosphate kinase concentration were determined by titration.
IModifications of the reaction components are detailed in descriptions of individual experiments.
The decrease in absorbance at 340 nm was measured with a Gilford model 2400 recording spectrophotometer thermostated at 25". This assay gives a 2-fold amplification of the rate, since 2 moles of NADH are oxidized per mole of n-fructose l-phosphate phosphorylated.
Since the molar absorptivity of NADH at 340 nm is 6.22 x lo3 M-l cm-l, 1 absorbance unit was equivalent to 12 nmoles of n-fructose 1,6-diphosphate formed.
Most rates were in the range of 0.1 to 0.6 nmoles of D-fructose l-phosphate phosphorylated per min. The manufacturer's specifications of the model 2400 recording spectrophotometer are given as a precision  Table  I shows the effect of various monovalent cations on the reaction catalyzed by D-fructose-l-P kinase. K+, Rb+, and NH,+ activated to different degrees, with K+ being the most effective. Li+ inhibited, and Na+ and Cs+ had little or no effect on the activity.
The activating effects of KCI, RbCl, and NH&I could not have been due to Cl-, since NaCl, CsCl (Table  I), and increased levels of MgClz (3) did not stimulate the reaction. The enzyme exhibited the same level of activity when NaOH was used instead of (CH3)4NOH to neutralize the reagents used for the assay. present at high concentrations, became greater when the concentration of either substrate was lowered, and was maximal when both substrates were present in small amounts (Table II). Fig. 3 shows graphically the effect of 40 mu KC1 on the n-fructose-1-P kinase reaction in the presence of various concentrations of XTP and n-fructose-l-P.
The theoretical curves were generated from the Michaelis equation for enzymic reactions involving two substrates in which neither substrate affects the K, for the other (18). The experimental points coincided well with the theoretical values. Activation by K+ in the presence of 3 mu ATP and n-fructose-l-P was about 4-fold, became greater as the concentration of both substrates was lowered, and was about lo-fold when both ATP and n-fructose-l-P were present at concentrations near their K, values. KA for K+--The apparent KA for K+ was dependent on the concentration of D-fructose-l-P (Fig. 4), being about 7 mM at 0.63 ITIM n-fructose-l-P and decreasing to about 3 mM when the Issue of June 25, 1970 V. Xapico and R. L. Anderson 3255 n-fructose-l-P concentration was increased to 12.6 mM. On the other hand, the apparent K, for K+ was independent of the concentration of ATP and Mg++ (Fig. 5).
Ej,fect of K+ on the Inhibition of o-Fructose-I-P Kinase Activity by d TP-ATP inhibited n-fructose-l-l' kinase activity when Mg++ to ATP ratios were less than 2: 1 (3). Although 40 mM KC1 did not totally overcome AT1 inhibition, a marked activation by K+ was observed when Mg++ to ATP ratios were less than 2: 1 (Fig. 6). At 6.6 InM ATP and 3.3 mM Mg++, 40 mM KC1 activated the enzyme over 100.fold, whereas at 6.6 mM ATP and 13.2 InM Mg++, the activation by the same concentration of KC1 was only about 4-fold. DISCUSSION In many cases of K+ activation, an absolute requirement for a monovalent cation has been demonstrated (7-9, 13, 19-21). In contrast, A. aerogenes n-fructose-l-P kinase, although activated, does not seem to require K+ absolutely for activity; removal of possibly activating monovalent cations from the reaction components prior to assay did not result in a total loss of activity, and the tetramethylammonium ion used in the assay has been reported to be nonactivating for enzymes affected by monovalent cations (8,21,22).' Na+ had no effect since n-fructose-l-P kinase activity was identical when either (CH,),NOH or NaOH was used to neutralize the assay reagents, and addition of NaCl to the assay containing (CHB)4N+ had no significant effect on the reaction velocity.
Like many other enzymes activated by K+ (7), n-fructose-l-P kinase was partially activated by Rb+ and NH4+ and was inhibited by Li+.
K+ affected not only the maximal velocity of the D-fructose-l-P kinase reaction, but also the K, for both ATP and n-fructose-l-P. The greater percentage activation observed at lower substrate concentrations is to be expected because of the ability of Kf to increase the apparent affinity of the enzyme for both substrates. In contrast to the finding with acetyl CoA synthetase (8), K+ activation of n-fructose-l-P kinase was not modified by excess Mg++ (Fig. 5). Inhibition of activity by ATP could not be relieved by n-fructose-l-P, ADP, AMP, or various other nucleoside mono-and diphosphates (3). However, K+ was effective in partially overcoming ATP inhibition when the Mg++ concentration was less than twice that of ATP (Fig. 6), although the requirement of the reaction for Mg++ could not be replaced by K+. NH,+ has been reported to relieve ATP inhibition of rat liver n-fructose-6-I' kinase (15), but not that from rabbit muscle (11,16).
It is apparent that monovalent cations affect different enzymes in a variety of ways. For example, yeast pyruvate kinase, which has an absolute requirement for monovalent cations, displays a sigmoidal rat,e-monovalent cation relationship (21), and K+ serves as an allosteric activator for muscle AMP deaminase, but has no effect on the maximal velocity of the reaction (22).
The simplest model which would account for our data may be represented as : where E represents n-fructose-l-P kinase and E-K+ represents an enzyme-K+ complex. E, which would predominate in the 1 Since the preparation of this manuscript, Faloona and Srere (23) reported that Escherichia coli citrate synthase activity is increased by tetramethylammonium chloride, but this may be an ionic strength effect rather than a specific activation. absence of K+, has K, values of about 0.7 to 0.8 MM for both ATP and n-fructose-l-P, whereas E-K+, which would predominate in the presence of excess K+, has K, values of about 0.3 mM for both substrates.
E-K+ exhibits about 3 to 10 times the activity of E, depending on the concentrations of n-fructose-l-P and ATP.
The binding of either ATP or n-fructose-l-P to the enzyme is not affected by the concentration of the other substrate (3). Therefore, if this model is correct, the following version of the Michaelis equation, as derived for a two-substrate reaction (18), should apply: where v is the velocity; [SJ and [&I are the concentrations of D-fructose-l-P and ATP, respectively; Ks, and Ksz are the apparent K, values for n-fructose-l-P and ATP, respectively; and k and e are constants which reflect the activity and amount of the enzyme.
Using experimentally determined values of Ks,, Ks,, and k for both the E and E-K+ forms of the enzyme, the calculated velocities did, in fact, coincide well with the observed velocities (Fig. 3). The variance of the KA for K+ with the D-fructose-l-P concentration (Fig. 4) may be explained by assuming the following reactions: where S,-E and X1-E-K+ represent complexes of n-fructose-l-P with E and E-K+, respectively.
The K, for K+ would tend to equal kz/kl in the presence of limiting n-fructose-l-P, and k4/k3 in the presence of excess n-fructose-l-P.
The data in Fig. 6 may be interpreted to mean that E-K+ is less susceptible to inhibition by excess ,4Tl' than is E. It should be emphasized that this model is consistent with K+ functioning either by interacting with substrate at the catalytic site or by inducing conformational changes through binding at an allosteric site. Although cooperative kinetics was not observed (3), it should be noted that allosteric modification of enzymes which exhibit hyperbolic ratesubstrate relationships have been previously described (24)(25)(26), and such cases have been treated theoretically by Koshland (27). Since the proposed model does not necessarily require the bindmg of K+ at the catalytic site, the mechanism of K+ activation of D-fructose-l-P kinase could be an exception to the generalized mechanism of monovalent cation activation of phosphoryl-transfer enzymes suggested by Suelter (28).
n-Fructose-6-P, n-fructose-l, 6-di-P, and citrate inhibit n-fructose-l-P kinase competitively with n-fructose-l-P (3). Such inhibition suggests possible control of activity in vivo. In view of the ability of K+ to overcome ATP inhibition partially, it would be of interest to determine the effect of this cation on inhibition by the above compounds. It is not known whether the n-fructose-l-P kinase-catalyzed reaction is a rate-controlling step in n-fructose metabolism in d. aerogenes. If it is, then it is likely that inhibition of n-fructose-l-P kinase by ATP, n-fructose-6-P, n-fructose-l, 6-di-P, and citrate, and activation by K+, play significant roles in the regulation of n-fructose metabolism, In Potassium Ion Activation of D-Fructose I-Phosphate Kinase Vol. 245, No. 12 any event, the specific influx of K+ that is known to accompany the uptake and metabolism of sugars by microbial cells (4-6) would tend to keep D-fructose-l-P kinase in a maximally active state.