Size-dependent Allosteric Effects of Monovalent Cations on Rabbit Liver Fructose-1,6-bisphosphatase”

Effects of monovalent cations on the neutral rabbit liver fructose-1,6-bisphosphatase are multifunc-tional and dependent on their nonhydrated ionic size. (a) The maximal velocity is increased by addition of monovalent cations with the optimum stimulation occurring with a nonhydrated ionic radius of 1.2 A in the presence of a chelating agent such as EDTA. (b) Activation curves are sigmoidal with n values varying from 1.5 to 2.3 as ionic radius of monovalent cation increases. The apparent K, values from 16.0 to 180 mM, obtained for various monovalent cations, have a linear relationship to ionic radii of cations. (c) At lower concentrations of fructose 1,6-bisphosphate monovalent cations show the inhibitory effect and the apparent K, for fructose 1,6-bisphosphate is increased as the concentration of monovalent cation is increased. A linear relationship is obtained between the slopes of increase in the K, and the reciprocals of ionic volume of monovalent cations. (d) The apparent K, for Mg *+ is also increased as the concentration of monovalent cation is increased, and a linear relationship is obtained again between the increases in K. and the reciprocals of ionic volume of monovalent cations. The cooperative nature for Mg*+ saturation is decreased as the K, increases. (e) The apparent Ki for AMP is also linearly altered as the concentration of monovalent cation is varied. However, the alteration of the K,

Effects of monovalent cations on the neutral rabbit liver fructose-1,6-bisphosphatase are multifunctional and dependent on their nonhydrated ionic size. (a) The maximal velocity is increased by addition of monovalent cations with the optimum stimulation occurring with a nonhydrated ionic radius of 1. the apparent K, for fructose 1,6-bisphosphate is increased as the concentration of monovalent cation is increased. A linear relationship is obtained between the slopes of increase in the K, and the reciprocals of ionic volume of monovalent cations. (d) The apparent K, for Mg *+ is also increased as the concentration of monovalent cation is increased, and a linear relationship is obtained again between the increases in K. and the reciprocals of ionic volume of monovalent cations. The cooperative nature for Mg*+ saturation is decreased as the K, increases. (e) The apparent Ki for AMP is also linearly altered as the concentration of monovalent cation is varied. However, the alteration of the K, is unusual, that is, the smaller cations than K+ increase the K, (Li+ > Na+ > NH,+), whereas the larger cations decrease the value ((CH,CH,OH),N+ > Cs+ > Rb'). The effect of K + is insignificant.
Alterations in the K, are also linearly related to the reciprocals of ionic volume of monovalent cations. The cooperative nature for AMP inhibition is decreased or increased as the K, increased or decreased. (f) In the absence of the chelating agent, the curves for Mg2+ saturation and AMP inhibition were hyperbolic without monovalent cations. By addition of monovalent cation the K, for Mg*+ or K, for AMP is increased and cooperative natures for binding of both ligands are induced. For nonspherical monovalent cations, the application of "functional ionic radius" is proposed. Functional ionic radii of NH,+, (CH,OH),CNH,+, and (CH,CH,OH),N+ are estimated to be 1.17, 2.55, and 2.87 A, respectively.
The presence of two distinct sites for the actions of monovalent cations is suggested.
Fructose-1,6-bisphosphatase (D-fructose-1,6-bisphosphate lphosphohydrolase, EC 3.1.3.11) catalyzes Reaction 1 in the presence of a divalent cation Mg'+ (1) or Mn*+ (2). D-Fructose-1,6-P.' + H,O -D-fructose-6-P + P, (1) Hubert et al. (3) reported that this enzyme is activated by monovalent cations such as K+ or NH,+. A large number of other enzymes have also been reported to be activated by monovalent cations (4)(5)(6). Regarding the action of monovalent cations, Eisenman (7) has predicted that the key to the cation selectivity is the anionic field strength of the binding site, and Melchoir (8), Lowenstein (9), and Suelter (5) suggest that the monovalent cation acts as a bridge between the enzyme and the substrate. McClure et al. (10) observed that not only anionic field strength but also nonhydrated ionic size of the monovalent cation is an important factor in activation of rat liver pyruvate carboxylase. On the other hand, Evans and Sorger (4) proposed that monovalent cations maintain a protein conformation necessary for optimum catalytic efficiency.

4315
*This work was supported in part by Scientific Research Funds 947112 and 967019 from the Ministry of Education of Japan and by a fund from the Foundation for Studies for Drug Resources (Japan).

4316
Size-dependent Effects of Monovalent Cations   To clarify  the mode  of action  of monovalent  cations  on  neutral  rabbit  liver  fructose-1,6-bisphosphatase,  the kinetic  parameters  of the enzyme  in the presence  of cations  are  investigated  and the sequence  specificity  of monovalent  cations is analyzed.  Evidence  is presented  for size-dependent  actions  of cations  on two distinct  sites of enzyme. EXPERIMENTAL PROCEDURE Materials-Glucose-6-P isomerase and glucose-6-P dehydrogenase were obtained from Boehringer Mannheim and mixed. Before use, these enzyme preparations were dialyzed against a thousand volumes of 10 rnM triethanolamine-HCl, pH 7.4, for 18 hours at 4". NaJructose-1,6-P,, NaNADP+, and NaAMP were purchased from Sigma. P-cellulose (Pll) was obtained from Whatman and washed with 1 N NaOH and HCl, and then extensively with water. All other chemicals were reagent grade.
Methods-Neutral fructose-1,6-bisphosphatase was purified from fresh male rabbit liver according to the methods of Tashima et al. (16) and Traniello et al. (17) with the following modifications. The animal was killed by a blow on the head, and the liver was perfused with cold 0.25 M sucrose containing 10 mM Tris-HCl, pH 7.4, before removal, and was chilled in the same solution.
The liver was chopped into small pieces and homogenized with 3 volumes (v/w) of the perfusion medium at 4' in a Potter-Elvehjem type homogenizer fitted with a Teflon pestle. The homogenate was centrifuged at 12,000 x g for 20 min, and to each liter of the supernatant fraction 243 g of (NH,),SO, was added slowly. After the precipitate was removed by centrifugation at 12,000 x g for 20 min, the solution (400 ml) was dialyzed several times against 5 liters of 0.1 mM EDTA, pH 7.0, at 4". The pH of the solution was adjusted to 6.5 by addition of 5 N acetic acid, and the P-cellulose paste corresponding to 3 g of dry powder was added, keeping the pH constant with 2 N NaOH. The P-cellulose slurry was removed by filtration on a Biichner funnel, and the filtrate was adjusted to pH 6.1 with 5 N acetic acid. Additional P-cellulose paste corresponding to 2 g of dry powder was added to adsorb fructose-1,6-bisphosphatase.
During the adsorption the pH was kept constant with 2 N NaOH. The suspension was filtered, and P-cellulose was packed in a column (2.5 x 20 cm) and washed with 10 liters of 0.23 M acetate buffer, pH 6.3, containing 0.1 mM EDTA. Fructose-1,6-bisphosphatase was eluted with the solution of 2 mM fructose-1,6-P, in the same buffer. The enzyme was precipitated and kept in (NH,),SO, solution of 80% saturation. For experiments the enzyme suspension was diluted to the concentration of 1 mg per ml with 10 mM triethanolamine-HCI buffer, pH 8.0, and was dialyzed against a thousand volumes of the same buffer for 18 hours at 4". Dialyzed enzyme preparation was diluted with 50 mM triethanolamine-HCl, pH 7.4, to desired concentrations before use. Fructose-1,6-bisphosphatase activity was assayed spectrophotometrically by following the formation of NADPH at 30" in the presence of excess glucose-6-P isomerase and glucose-6-P dehydrogenase. The standard assay system (1 ml) contained 0.1 mMfructose-1,6-P,, 2.5 mM MgCl,, 0.1 mM EDTA, 0.2 mM NADP+, 7 units of glucose-6-P isomerase, 5 units of glucose-6-P dehydrogenase, and 40 mM triethanolamine-HCI, pH 7.4. Glucose-6-P isomerase and glucose-6-P dehydrogenase were added before the addition of fructose-1,6-bisphosphatase to the cuvettes, in which other reagents including monovalent cation had been mixed well. All reagents for enzyme assay were adjusted to the neutral pH with NaOH or HCl when necessary.

Activation by Monovalent
Cation-Monovalent cation acti-fructose-1,6-P, higher than 0.025 mM (Fig. 1). With respect to Li+, activation was not observed at any concentration. The activity in the absence of alkali metal ions or NH,+ corresponds to that in the presence of 40 mM triethanolamine alone. The enzyme showed a partial dependency on monovalent cation, and at lower concentrations of triethanolamine than 25 mM, enzyme activity became constant to approximately 40% of that observed in the presence of saturating NH,+. Moreover, activation by 40 mM triethanolamine, which was used for the buffer solution in the studies of the effects of monovalent cations, was almost negligible (less than 3%). Thus it was practicable to analyze the kinetic parameters in the presence of added monovalent cations in this buffer. Activation curves were found to be sigmoidal, and this cooperative nature was increased as ionic size of the monovalent cation increased. The estimated n values for Na+, NH,+, K+, Rb+, Cs+, Tris, and triethanolamine were 1.5, 1.5, 1.6, 1.6, 1.7, 2.1, and 2.3, respectively.
Although activation was observed with potassium salt of any anion tested, the sulfate salt was most effective as was previously observed with kidney enzyme (3), whereas phosphate salt diminished the activating effect by about 40%. In the following experiments, sulfate salts of monovalent cations were used throughout.
These activating effects of monovalent cations were not evident in the absence of a chelating agent such as EDTA (Table I). Unless otherwise noted, 0.1 mM EDTA was added in the assay system throughout the experiments.
K, and Ionic Radius-Apparent activation constants (K.s) for monovalent cations obtained from the concentrations for 50% activation in Fig. 1   Functional nonhydrated ionic radii of Tris and triethanolamine were also estimated to be 2.55 and 2.87 A, respectively, from their K, values.

Maximal Activation
and Ionic Radius-When the maximal velocities in the presence of various monovalent cations were plotted against ionic radii of cations, optimal ionic radius for the maximal activation was observed at about 1.2 A (Fig. 2B).
Fructose-1,6-P, Saturation and Monovalent Cation-The saturation kinetics of the enzyme for fructose-1,6-P, was affected by the presence of a monovalent cation, and at lower concentrations of fructose-1,6-P, the monovalent cation showed an inhibitory effect (Fig. 3). The Michaelis constant for fructose-1,6-P,, which was estimated from the data at concentrations lower than 0.02 mM, was linearly increased as the concentration of monovalent cation increased (Fig. 4). The K, in the absence of alkali metal ions or NH,+ corresponds to the value of 6.1 x 1OmB M for 40 mM triethanolamine alone which is present as the buffer. This value is similar to the values for neutral kidney enzyme of 5 to 10 x 10 -' M in the presence of 40 mu triethanolamine and 40 mM diethanolamine (16). From  Fig. 4, the K, of neutral rabbit liver enzyme for fructose-1,6-PZ, which would be obtained in the absence of monovalent cations, was estimated to be 4.0 x 1OmB M.

K, and Ionic
Volume-Increase in the K, was linearly correlated with the reciprocal of ionic volume of monovalent cation (Fig. 4, inset). Ionic volumes obtained for NH,+ and (CH,CH,OH)sN+ from ,the data of Fig. 2A are also in good correlation to the increases in the K, value.
These effects of monovalent cations on the K, were also observed in the absence of a chelating agent.

Mg'+ Saturation and Monovalent
Cation-Monovalent cations altered the affinity of neutral rabbit liver fructose-1,6-bisphosphatase for the cofactor divalent cation, MgZ+, increasing its apparent K, value obtained as the concentration for 50% activation. Fig. 5 shows the effect of Na+ on Mg2+ saturation curves in the presence of 0.1 mM EDTA. The curves were sigmoidal and the increased K, values were obtained by addition of Na+ as previously reported for 150 mM K+ (3).
Similar effects were observed with other monovalent cations, and the K, for Mg '+ linearly increased as the concentration of each monovalent cation increased (Fig. 6). The K, in 40 mM triethanolamine buffer alone was 0.20 mM, and the basic K, for MZ+> which would be obtained in the absence of any monovalent cation, was estimated to be 0.16 mM.
The sigmoidal nature of MgZ+ saturation curves for other fructose-1,6-bisphosphatases had been reported previously (3). Neutral rabbit liver enzyme also showed the cooperative nature in MgZ+ saturation with the n value of 1.95 in 40 mM triethanolamine, and this value was decreased in the presence of Li+ or Na+. With 30 mM Li+ and 300 mM Na+, the n values of 0.33 and 1.12 were obtained, respectively. On addition of larger monovalent cations, decrease in the value was very slight.

K, for Mg2+
and Ionic Volume-The increase in the K, for Mg2+ was also found to be related to ionic volume of monovalent cation. A linear relationship was observed between the rates of increase in the K, and the reciprocals of ionic volume of monovalent cations (Fig. 7).
Effect of EDTA-The K, and n values for Mg2+ saturation were influenced by removal of the chelator EDTA. In 40 IrIM triethanolamine buffer containing no EDTA, the Mg2+ saturation curve was found to be almost hyperbolic showing the n value and K, of 1.02 and 0.42 mM, respectively (Fig. 8). High concentrations of the monovalent cation, however, induced the sigmoidal nature of the MgZ+ saturation curve in the absence of EDTA, and with 150 IIIM K+ the n and K, values were increased to 1.25 and 2.1 mM, respectively. Thus in the absence of EDTA, increase in the K, for Mg2+ by addition of monovalent cation was more marked, whereas increase in the maximal velocity was very slight.

AMP Inhibition and Monovalent
Cation-Effect of monovalent cation on the affinity of rabbit liver fructose-1,6-bisphos- phatase for AMP was interesting. Larger cations than K+ increased the affinity, whereas smaller cations decreased it (Fig. 9). The n for AMP inhibition of 1.76 was also altered slightly in the presence of high concentrations of monovalent cations. Larger cations than K+ increased the value and smaller ones decreased it.
The apparent Kt value for AMP was also linearly altered as the concentration of monovalent cations increased (Fig. 10). K+ exhibited almost no effect on the K, value at any concentration.
This may be the reason why Hubert et al. the absence of monovalent cations was estimated to be 1.32 x lk5 M. These effects of monovalent cations on AMP inhibition were observed in the absence of the chelator EDTA, too. AMP inhibition curve in the absence of EDTA was almost hyperbolic, and by the addition of 150 mM Na+ the Kr and n values were altered from 1.66 x 10m5 M and 1.1 to 4.37 x 1O-6 M and 1.4, respectively. That is, the monovalent cation in the absence of a chelator induces the cooperativity for AMP inhibition, as well as for Mg*+ saturation.
KI and Ionic Volume-When the slopes of alteration in the K, were plotted against the reciprocals of ionic volume of monovalent cations, a linear relationship was obtained again as shown in Fig. 11. Critical ionic volume, which would cause no alteration in the K,, was estimated to be 9.80 A" corresponding to ionic radius of 1.33 A.

DISCUSSION
The kinetic constants for all ligands, namely, the K, for fructose-1,6-P,, K, for Mg2+, and Kt for AMP, were increased or altered as the concentration of monovalent cations increased. As has been discussed previously by McGregor et al. (23), a general conformational change in enzyme protein may result in the alteration of the constants for all substances. One of the effects of a monovalent cation on rabbit liver fructose-1,6-bisphosphatase may be a general conformational change in enzyme protein. Linear relations are observed between the rates of alteration in these constants and the reciprocals of ionic volumes of monovalent cations (Figs. 4, 7, and 11). A possible explanation of the mechanism for these phenomena would be that the extent of conformational change is dependent on the amount of ions which could get into a specific "pocket," distinct from the catalytic site or AMP binding site. The alteration in the Ki for AMP is, however, different from that for the K, for fructose-1,6-P* or the K,, for Mg 7 2+ that is K+ shows the neutral effect and smaller cations than K+ increase the constant whereas larger ones decrease the value. Probably a general conformational change may give different effects on different loci of enzyme protein. Hill plots of the data are obtained using the K, and n values estimated as described under "Experimental Procedure" and are shown in the inset. The maximal velocities were obtained as described in the legend for Fig. 5.
The cooperativities which are observed for Mg2+ activation and AMP inhibition in the presence of a chelating agent EDTA are decreased or increased as the K, and K, increased or decreased, respectively, by addition of monovalent cations. In the absence of the chelator, however, Mg'+ activation and AMP inhibition are noncooperative and the monovalent cation induces the cooperativity for both cases. All these observations indicate that the monovalent cation acts as a potent allosteric effector on rabbit liver fructose-1,6-bisphosphatase.
Another role of the monovalent cation in the regulation of fructose-1,6-bisphosphatase activity may be the direct participation in enzymatic catalysis at the catalytic site. The increase in the maximal velocity is dependent on ionic radius of the monovalent cation with the optimum radius of 1.2 A. It is of interest that this activation by a monovalent cation requires the presence of a chelator such as EDTA (Table I) whereas the conformational effect of the cation mentioned above can occur even in the absence of the chelator. These results suggest that in the presence of a chelating agent the monovalent cation is also participating in the direct catalysis as a bridge between ligands, or between enzyme and a ligand as has been suggested in other enzyme reactions (8)(9)(10)(23)(24)(25) The inset shows the data for Li+. The lines were drawn using the slopes and intercepts obtained as described in the legend for Fig. 4. FIG. 7 (right). Relationship between the increase in the K. for Mgz+ and ionic volume of monovalent cation. The slopes of Fig. 6 are replotted against the reciprocals of ionic volume of monovalent cations. Ionic radii of NH,+, Tris, and triethanolamine were obtained from Fig. 2A. pyridoxal Y-phosphate, cysteine, P-mercaptoethanol, or oleate on fructose-1,6-bisphosphatase have been extensively investigated and discussed. It is now established that activation of fructose-1,6-bisphosphatase by a chelator is not due to the removal of trace heavy metal contaminants (26)(27)(28)(29)(30)(31). The activation mechanisms suggested so far are: (a) the direct involvement in catalysis probably in such a chelate form as Mg*+EDTA (28)(29)(30)(31)(32) which would be acting on a different form of enzyme from that for free Mg'+ (29-31), (5) prevention or removal of ATP inactivation (33,34), and (c) of GSSG inactivation (35). We now present evidence for additional actions of the chelating agent, that is, (d) participation in monovalent cation activation and (e) induction of cooperativities in Mg2+ saturation and AMP inhibition in the absence of monovalent cations.
In the activation reaction of rabbit liver fructose-1,6-bisphosphatase, saturation curves for monovalent cations are sigmoidal showing that the binding of a monovalent cation to the catalytic site is also cooperative (Fig. l), while hyperbolic saturations are generally observed in other enzymes (6,23,36).
The extent of cooperativity and the apparent activation constant possess the tight relationships to ionic radius of the monovalent cation. When the previously reported ionic radius ' is employed, NH,+ deviates from the linear relationship. Non-metal cations such as NH,+, (CH,OH),CNH,+, or (CH,CH,OH),N+ are not completely spherical, so that their effective ionic radii or ionic volumes in the biological system may differ from the physically reported values. Using fructose-1,6-bisphosphatase activating system, it may be possible to obtain the "functional" radii of various monovalent cations. Thus the presence of two distinct sites for the actions of the monovalent cation is suggested. One is the catalytic site in which the monovalent cation is promoting the enzymatic catalysis. Another is the allosteric site which is involved in the alteration of kinetic constants for the substrate and effecters indicating a general conformational change. The action of the monovalent cation on the former site can be distinguished from that on the latter by (a) the requirement for a chelating agent, (b) the cooperative nature for its binding, and (c) the correlation to ionic radius of the monovalent cation whereas ionic volume is an important factor for the latter.
Similarly, two distinct effects of monovalent cation are recently observed in the studies on rat liver pyruvate carboxylase. A monovalent cation is required for its activity (lo), while high concentrations of the cation affect the enzyme conformation and cause the dissociation of the native tetrameric enzyme into enzymatically active dimers or protomers, and finally into inactive protomers (37). With rabbit liver fructose-1,6-bisphosphatase, however, no evidence for the dissociation of enzyme protein in the presence of a monovalent cation has been obtained.
Initial velocity pattern for sat,uration of fructose-1,6-P, in the presence of monovalent cation (Fig. 3) is complex and shows the characteristic of double competitive substrate inhibition which has been first observed in yeast j%ketothiolase (38) and is characterized by Cleland (39). However, to clarify whether monovalent cation is acting as a co-substrate in the reaction of fructose-1,6-bisphosphatase, further examinations are needed.
Considering that K+ is the principal constituent of intracellular monovalent cations with the concentration reaching up to 150 meq per liter (40), it may be physiologically significant that K+ is one of the best activating monovalent cations for fructose-1,6-bisphosphatase with the optimum concentration of 150 mM. The increased K, for MgZ+ in the presence of high concentrations of monovalent cations is also as described under "Experimental Procedure." V, and u are the uninhibited and inhibited rates, respectively. FIG. 10 (right).
Effect of monovalent cation on the K, for AMP. The apparent K, values, obtained as described under "Experimental Procedure," were plotted against the concentration of monovalent cations. The lines were drawn using the slopes and intercepts estimated as described in the legend for Fig. 4. The data for Lit are in the inset. consistent with the very high intracellular concentration of MgZ+ of around 30 meq per liter (40). Conversely, it is of interest that K+ does not interfere in the regulatory action of the allosteric inhibitor AMP (20, 41) on liver fructose-1,6bisphosphatase.