Activation of Yeast Pyruvate Kinase by Natural and Artificial Cryoprotectants*

SUMMARY The inactivation of yeast pyruvate kinase at temperatures near 0’ and the stabilization achieved by addition of 50% glycerol or 0.1 M KC1 and 24 mu MgClz recently documented prompted an examination of the effects of glycerol and other cryoprotectants on the kinetic activity of this enzyme. We now report that addition of glycerol, dimethyl sulfoxide, trimethylphosphine oxide, or dextran to an assay mixture containing less than K, levels of substrates activates yeast pyruvate kinase often to the same extent as that observed by addition of the positive modifier, fructose 1,6-diphosphate, in the absence of the cryoprotectant. Maximum effects were observed with 33% glycerol, 20% dimethyl s&oxide, and 14% trimethylphosphine oxide. The activation by glycerol and dimethyl s&oxide at 30” results from a lowering of the K, for phosphoenolpyruvate from 1.1 mu to 0.3 to 0.4 mM and the Hill slope (nrr) from 3.1 to 2.1.

The inactivation of yeast pyruvate kinase at temperatures near 0' and the stabilization achieved by addition of 50% glycerol or 0.1 M KC1 and 24 mu MgClz recently documented (Kuczenski, R. T., and Suelter, C. H., Biochemisfry, 9, 939, 1970) prompted an examination of the effects of glycerol and other cryoprotectants on the kinetic activity of this enzyme. We now report that addition of glycerol, dimethyl sulfoxide, trimethylphosphine oxide, or dextran to an assay mixture containing less than K, levels of substrates activates yeast pyruvate kinase often to the same extent as that observed by addition of the positive modifier, fructose 1,6-diphosphate, in the absence of the cryoprotectant.
Maximum effects were observed with 33% glycerol, 20% dimethyl s&oxide, and 14% trimethylphosphine oxide. The activation by glycerol and dimethyl s&oxide at 30" results from a lowering of the K, for phosphoenolpyruvate from 1.1 mu to 0.3 to 0.4 mM and the Hill slope (nrr) from 3.1 to 2.1. Kinetic analysis at 0" in the absence of the cryoprotectant gave a K, of 0.13 mM and nH = 0.94. These changes in the kinetics of yeast pyruvate kinase are discussed in terms of the effect of temperature and solvent on the structure of both water and enzyme.
Yeast pyruvate kinase because of its cold lability (I) must be stabilized during purification at 4". Glycerol, a natural cryoprotectant, which was used for this purpose and which is also rtqllired for renaturation of guanidine HCl denatured enzyme' also protects many biological materials against freeze-thaw danlage (2). Other compounds such as dimcthyl sulfoxide have been used extensively in low temperature work (a), and dextran has been shown to protect rat hearts (3), erythrocytes (4, 5), aud bone marrow (6) from extreme freeze-thaw damage. properties for red blood cells, but had not yet been tested (7). This paper reports the results of our study of the effects of these compounds on the catalytic activity of yeast pyruvate kinase examined at 30 and 0".

MATERIALS
Yeast pyruvate kinase was prepared by the method of Hunsley and Suelter (1). Ammonium sulfate suspensions of rabbit muscle lactic dehydrogenase, Na2ADP, tricyclohexylammonium-P-enolpyruvate,2 and tetracyclohexylammonium-fructose-l ,6di-1' were obtained from Sigma. The Grade A dithiothreitol was purchased from Calbiochem.
Dimethyl sulfoxide (Aldrich, industrial grade) was shaken with NaHCOa and distilled in vacw, at approximately 90". Between experiments it was stored frozen. Dextran was purchased from K and K Laboratories, Inc., Plainview, New York (average mol mt 86,000).
Reagent grade glycerol was a product of J. T. Baker Chemical Company, Phillipsburg, New Jersey. Trimethylphosphine oxide was synthesized by the method of Burg and McKee (8) with a bromide Grignard instead of the reported chloride reagent.
The compound melted at 138 to 139' after vacuum sublimation.
A Beckman DB spectrophotometer was used for all assays except those which cont,ained trimethylphosphine oxide as part of the assay mixture.
For these a Beckman DU with a Gilford automatic sample changer attachment was used.

METHODS
Yeast pyruvate kinase was desalted by passing a solution of the enzyme over a column of Sephadex G-25 (coarse, 0.8 x 11.0 cm) equilibrated with 0.1 M Tris, p1-I 7.5, and eluted with the same buffer. Aliquots were tested with saturated BaC12 solution to insure them free of ammonium sulfate.
Protein concentrations were estimated with the absorption coefficient Ey.AE = 0.653 at 280 nm (1). For the kinetic studies, the enzyme at 0.25 mg per ml was incubated for 3 hours at 23" in 0.1 11 Tris, pH 7.5, 0.23 nf KCl, 25 rnnf MgC12, and 50 mnf dithiothreitol to avoid the superactivation phenomena rcportcd by Kuczenski and Suelter (9). Dithiothreitol, K+ and Mg2+ were required to prevent loss in the catalytic activity of the enzyme when assayed at suboptimal concentrations of substrate.3 Yeast pyruvate kinase was assayed according to the method of Hunsley and Suelter (10) with the following modifications.

Issue of October 10, 1971
M. J. Ruwart and C. H. Xuelter 5991 Phosphoenolpyruvate was 0.4 mu; ADP, 5 mu; and MgC12, 17 mu, as compared with 5 mu, 10 mu, and 24 mM, respectively. The concentration of NHd+ introduced with the lactic dehydrogenase was calculated to be 12 mM, while the concentration of Na+ from the NADH was approximately 30 mu. In all cases the enzyme solution was maintained at 23" and the assays were run at 30" in a total volume of 1.00 ml, except in the assays containing (CHI)aPO where the volume was 0.25 ml. For assays at O", the temperature of the Beckman DB was monitored with a calibrated thermistor. The cuvette chamber was dehumidified with heat-activated silica gel (Sigma). The lactic dehydrogenase concentration was increased 7-fold to compensate for its lower catalytic rate at 0". RESULTS E$ect of CryoprotectuntsIn order to examine the effect of glycerol and other cryoprotectants on the kinetic parameters of yeast pyruvate kinase, the enzyme was assayed at suboptimal concentrations of substrates, particularly P-enolpyruvate, in the presence and absence of fructose-1,6-di-P at increasing concentrations of the test compounds. Suboptimal concentrations of substrates were chosen with the hope that either an activation or an inhibition would be readily discerned. Fig. 1A shows the effect of increasing concentrations of glycerol on the activity of the enzyme. The activation factor shown on the ordinate is the ratio of the catalytic activity observed under each condition divided by the catalytic activity obtained in the absence of fructose-l, 6-di-P and glycerol. The ratio determined with glycerol and fructose-l, 6-di-P in the assay increases slightly as glycerol is added reaching a peak at 2.3 M (16%) glycerol but then decreases as the concentration of glycerol is increased. Doubling the concentration of lactic dehydrogense in the linked assay did not change the results. At 4.9 M (33%) glycerol, the ratio and thus the catalytic activity without fruc-tose-1 ,6-di-P is nearly identical with that obtained with fructose-1,6-di-P in the absence of glycerol.
The effect of varying (CH&SO is presented in Fig. 1B. The observations are similar to those observed with glycerol but the peaks in the activation ratios were at, 2.8 M (20%) (CH&SO in the absence of fructose-1,6-di-P and 2.1 M (15%) (CH&SO in the presence of fructose-l, 6-di-P. The decreased ratio observed at high concentrations of (CH&SO is not due to inhibition of lactic dehydrogenase (11) since increasing lactic dehydrogenase had no effect. The reported interaction of (CH&SO with pyruvate to produce a 340-nm adsorbing material (11) was not observed.
The maximum activation ratio for (CHJrPO as the concentration of (CHs)rPO is increased in the assay in the absence of fructose-l ,6-di-P occurs at 1.4 M (14%) (CHS8PO; the activation ratio in the region of (CHs),PO concentrations at which a peak in the activation ratio was observed with fructose-l, 6-di-P plus other activators was not determined (Fig. 1C). The activation ratios in (CHa)aPO were not as great as those observed in (CH&SO and glycerol. Dextran also produced an activation ratio of 4 in the absence of fructose-l, 6-di-P which was still increasing at 37% w/v dextran, but because of viscosity effects, completion of the study was not possible. Phosphoenolpyruvate Kinetics-In an effort to understand the activation observed in the binary solvents, the saturation kinetics of P-enolpyruvate was examined under several conditions noted in Table I. In each case the apparent Km and Hill slope (nH) was obtained from a standard Hill plot (12) of log [(Vmax/v) -l] plotted versus log[P-enolpyruvate]. It was previously shown that P-enolpyruvate saturation kinetics were sigmoid with nE s 3, Km.wp G 1 mM; in the presence of fructose-l, 6-di-P, saturation was normally hyperbolic with nnH E! 1.0 and Km r! 0.1 mu (10, 13). Assaying at 0" (Table I) I  I  I  I  I  I  I  I  I  I  I  I  I  I  0 1 1. A, the effect of glycerol on the activity of yeast pyru-sayed with fructose-1,6-di-P; O-0, assayed without fructosevate kinase assayed at suboptimal concentrations of substrate as 1,6-di-P. B, the effect of (CH&SO on the activity of yeast pyrudescribed under "Methods." The data are presented in terms of vate kinase assayed at suboptimal concentrations of substrate as an activation factor which is the activity of the enzyme at the described under "Methods." C, the effect of (CHS)~PO on the indicated glycerol concentration divided by the activity obtained activity of yeast pyruvate kinase assayed at suboptimal concenin the absence of fructose-1,6-di-P at 0% glycerol. O-0, as-trations of substrate as described under "Methods." E'urthcr, inclusion of fructose-l, 6-di-1' in the assay at 0" has no effect on kinetics.
.Uthough the concentrations of glycerol and (CH&SO used in the kinetic studies gave similar activation ratios as fructose-l, 6di-P alone at 0.4 mM P-enolpyruvate, the same kinetic parameters for P-cnolpyruvatc kinetics were not observed in glycerol and (CI-I&SO as with fructose-1,6-di-P.
The nH values were decreased slightly but cooperativity was not abolished. However, the Ii, for P-enolpyruvate was lowered significantly when assayed in 33% glycerol and 20%;, dimethyl s&oxide.

DISCUSSION
Generally the cryoprotcctive solvents tested in this study produced the same qualitative effects as the normal physiological activator fructose-l, 6-di-P (10, 13) on the catalytic properties of yeast pyruvatc kinase (Table I). Further an interaction between the activation by fructose-l, 6-di-P and the cryoprotcctant was observed since a maximum in the activation is observed at lower concentrations when fructose-l ,6-di-P is included in t,he binary solvent.
It is noteworthy that the extent of activation observed was greatly dependent on P-enolpyruvatc concentration.
In 20% (CH&SO, for example, the activation ratio in the absence of fructose-l ,6-di-P went from 18 to 1.5 as Penolpyruvatc concentration was increased from 0.3 to 1.6 mM. Depending upon the sensitivity of the assay, a substance which activates at one substrate concentration may not appear to induce any rate enhancement at another. This behavior should be considered when potential activating substances are being screened. The nonspecific activation observed in the binary solvents, although not clearly understood, is likely due to a perturbation in the balance of the various forces important in defining the steady state three-dimensional structure of a protein molecule in solution. For example, the ability of (CHJ2S0 (14) as well as 2Hz0 (15) and sucrose (16) to mimic the effect of GTP on glutamic dehydrogcnase by promoting inactive monomers and reducing the concentration of GTP required to achieve half-maximal effects was ascribed to the particular ability of these compounds to form hydrogen bonds (17) thereby inhibiting association of nlonomcrs.
Association of monomers was also inhibited by partial acetylation of amino groups consistent with the sugges-X Polyvinylpyrrolidone tion that free amiiio groups are important in the association phenomenon (18). Another perturbation in the balance of forces dictating protein structure may result from an interaction of these compounds with the solvent water, per se. For example, polyvinylpyrrolidone, an effective cryoprotectant, was shown to enhance the struct'urc of water as evidenced by infrared spcctroscopy (19) and to deviate considerably from ideality in freezing point depression experiments, i.e. a much greater freezing point depression of water is observed than one would expect from its osmolarity (20,21). Physiologically analogous compounds, glycoproteins of molecular weight between 10,000 and 20,000 isolated and characterized by DeVrics, Komatsu, and Feeny (22) from antartic fishes, also depress the freezing point of water much more than one would calculate from the number of particles in solution.
This excess freezing point depressing activity which was destroyed by acctylation of three amino groups per 10,000 g of protein or by cleavage of a few peptide bonds was suggested to be due to its expanded structure and the presumed immobilization of water (22). Evidence for alterations in the structure and activity of water due to the addition of hydrogen-binding solvents has been observed not only by infrared spectroscopy and freezing point depression, but also by examination of a wide range of other physical processes such as solvation of ions and nonelectrolytes, hydrolytic reactions, oxidation-reduction reactions, and ionization of weak electrolytes (23,24). Consideration of these observations in light of the structured and nonstructured twostate model for water structure (25) suggests that addition of O.l-to 0.2~mole fraction of glycerol, (CH&SO or (CHJgP = 0, or lowering the temperature of water arc perturbations that participate in the water structure-making process.4 Enzymes, such as pyruvate kinase, which arc catalytically sensitive to allosteric cffectors, might then be expected to display altered catalyt,ic properties when assayed under other solvating conditions as used in this study. In this case, addition of the cryoprotectants allowed the formation of an enzyme configuration with a greater affinity for P-enolpyruvate; cooperativity in the P-enolpyruvatc saturation kinetics was still observed. The decrease in the activation factor as the concentration of the cryoprotcctant is increased beyond the 33% glycerol (0.12.mole fraction), 20%) (W&SO (0.06.mole fraction) or 14% (CH~SPO (0.03-mole fraction) may be due to a "falling apart" of the structure, a term previously suggested by hrnett and R'IcKelvy (26) * Structure-making and structure-breaking processes are operationally defined tern& used in discussion of water solutions: AS suggested bv Lumrv and Rajender (23) they may be a manifestatill of the" enthalhy-entropy compensation law. The thermal denaturation of a variety of proteins which are related by a simple linear equation, AS+ = a AH+ + b, also constitutes a compensation law behavior (23).