Crystalline Pyruvate Oxidase from Escherichia coli

Crystalline pyruvate oxidase, a soluble tetrameric flavoprotein from Escherichia coli, which binds both thiamine pyrophosphate (TPP) and FAD is activated 15to loo-fold by phospholipids and long chain fatty acids. Maximal activation of the oxidase requires incubation of the enzyme for at least 6 min with the lipid activator in the presence of substrate and cofactors (pyruvate, TPP, and MgC12). Very little activation occurs if any of these components are omitted from the mixture. Furthermore, activation is markedly reduced if the enzyme is incubated with phospholipid before pyruvate, TPP, and MgCL are added. Phosphatides dramatically effect the kinetic parameters of the pyruvate oxidase reaction. The K,,, values for pyruvate and TPP are lowered 13-fold and 3to 4-fold, respectively, in the presence of phospholipid. In addition, phosphatides bestow cooperativity to the enzyme with respect to binding of TPP in a very unusual fashion. In the absence of phospholipid, TPP binding to the enzyme follows ordinary Michaelis-Menten type saturation kinetics. In the presence of phospholipid, TPP binds cooperatively to the enzyme and shifts the K,,, for TPP to a lower value. Stopped-flow experiments measuring the rate of reduction of enzyme-bound FAD clearly indicate that the presence of lipid activator affects a rate-controlling step leading to the formation of enzyme-FADHI. The rate of reduction of enzyme-bound FAD is increased at least lOO-fold in the presence of phospholipid. The physiological significance of pyruvate oxidase and the activation phenomenon are discussed.

:liuleriaZs--The materials used for the experiments described in this paper were obtained from the sources listed in the previous paper (1).
2,6-lXchlmoindophe7wl Reductase Assay-This assay procedure is based on the utilization of 2,6-dichloroindophenol as an electron acceptor for pyruvate oxidase.
The assay reaction mixture contained, in a total volume of 1 ml, 100 pmolcs of potassium phosphate buffer, pH 6.0, 10 pmoles of MgCl,, 50 pmoles of po- The order of addition of reactants to the DC1 assay mixture is important. The phospholipid activator is normally added after all other components except DC1 which was used to initiate the reaction. The assay mixture minus DC1 is incubated for 10 min before the addition of DCI. Unless otherwise noted, the phospholipid activator for the enzyme was a water-soluble micellar preparation of E. coli phosphatidylethanolamine (1). In most assays, 30 pg of phosphatidylethanolamine were added; however, smaller amounts (5 pg) of freshly prepared micelles of phosphatidylethanolamine are sufficient to fully activate the enzyme.
The reduction of DC1 is followed spectrophotometrically by measuring the loss of absorbance at 600 nm. DC1 solutions at a concentration of 2 X KY M were prepared in 0.1 M phosphate buffers ranging in pH from 6.2 to 7.9. The optical densities of these solutions were measured at 600 nm to establish the extinction coefficient (~600) as a function of pH (w---m).
To obtain the dependence of DC1 reductase activity on pH (O---O), a series of assay mixtures were prepared in tubes with the pH ranging from 6.2 to 7.8. DC1 solutions were prepared in the same buffers and added to their counterparts in the incubation mixture.
Assays could not be run below pH 6.2 because of the low solubility of DCI.
One unit of enzyme activity in the DC1 reductase assay is defined as the loss of 0.001 in absorbance per min.
Specific activity is defined as the number of enzyme units per mg of protein.
Purijication of Pyruvate Ox&se-Pyruvate oxidase was purified by the method of Williams and Hager (2) employing the modifications outlined in the preceding paper (1). Stopped-jlow ExperimentsThe rate of reduction of pyruvate oxidase with pyruvate was determined in a Gibson stopped-flow apparatus (3) by measuring loss of absorbance at 438 nm, the absorption maximum of the flavin group. The enzyme mixture, containing pyruvate oxidase, TPP, MgC12, and potassium phosphate buffer, pH 5.7, was placed in one syringe of the stoppedflow apparatus.
Lysolecithin (6.8 X 1O-4 M) was included in the enzyme mixture when measurements were made in the presence of phospholipid.
The other syringe contained 0.2 1w potassium pyruvate.
The contents from the syringes were mixed at a ratio of 1: 1, by volume, in a cell having a 2-cm light path. After mixing, the concentration of the enzyme, TPP, MgC12, and pyruvate were 1.7 X lo+ M, 1 X 10e3 M, 1 X 10m2 M, and 0.1 M, respectively.
The absorbance decrease (percentage of transmission increase) was recorded on a Tektronix type 564 storage oscilloscope fitted with a Polaroid camera. The rates measured were used to calculate the rate constants and half-life periods.

Conditions
for Maximal Activity of Enzyme-The oxidationreduction indicator, DCI, is an excellent electron acceptor in the pyruvate oxidase reaction. The K, for DC1 determined from the Lineweaver-Burk plot shown in Fig. 1 is 9.3 X low5 M. This dye is useful for assaying the enzyme at pH values as low as 6; however, at lower pH values DC1 is insoluble in concentrations required for the assay (2 x 10e4 M). As evidenced from Fig. 2, the pH optimum for DC1 reductase activity is 6 or lower.
This value is in agreement with the observed pH dependence of pyruvate oxidase when either ferricyanide or oxygen serve as terminal electron acceptors (2) Saturating amounts of micellar phospholipid were added to the mixture which was then allowed to incubate at room temperature for a predetermined time interval.
The reaction was then initiated by addition of the DC1 solution.
The time interval between addition of phospholipid and DC1 was varied from 15 set to 60 min.

Issue of March 25, 1971
C. C. Cunningham and L. P. Hager 1585 oxidase in several assays is dependent on the presence of a lipid activator.
Activation of the enzyme with phospholipid is a time-dependent process.
As shown in Fig. 3, about 6 min of incubation are required for complete activation of the enzyme by phospholipid.
The conditions for maximal activation of pyruvate oxidase have been considered in detail (see Table I).
All assays in Table  I  (Line A). In contrast, the activity of an enzyme preparation which had been incubated with phospholipid, substrate, and cofactors for 10 min was several-fold higher (Line B) .  Table II.
These values were obtained from the Lineweaver-Burk plots shown in Fig. 5  In addition, the presence of phosphatide resulted in a 12-fold increase in the V,,, for this particular enzyme sample.
The kinetics of the pyruvate oxidase reaction with respect to TPP are also altered dramatically in the presence of phospholipid.
The dependence of the velocity of the DC1 reductase activity as a function of the concentration of TPP in the presence and absence of phospholipid is shown in Fig. 6 (upper picture) and in the presence of lysolecithin (lower picture) are described under "Experimental Procedure." V max (6). In the case of TPP binding to pyruvate oxidase, phospholipids induce cooperativity while lowering the K, for TPP and raise the Vmax for the reaction by a factor of 12-to 25fold.
Hill plots (4) for TPP, shown in Fig. 7, establish that phospholipid changes the kinetic order of the reaction with respect to TPP. The "72" value obtained if phospholipid is present in the reaction mixture is 2.1. When no phosphatide is added to the reaction mixture, the slope of the Hill plot is 1 with respect to TPP. E$ect of Phospholipid on Rats of Reduction of Pyruvate Oxiclase by Pyruvate-To study the rate of reduction of pyruvate oxidase in the absence of added electron acceptors, a Gibson stoppedflow apparatus was employed (3). Reduction of the enzyme- The values below were obtained from the oscilloscope tracings shown in Fig. 8  bound FAD was measured both in the presence and absence of phospholipids.
One of the difficulties encountered in these experiments was the insolubility of the phospholipid micelles in the enzyme assay reaction mixture. All cephalin type phospholipid preparations precipitated immediately when added to the standard incubation mixture. Micellar lecithin was relatively stable in the assay solution in the absence of pyruvate; however, a precipitate formed immediately upon the addition of pyruvate.
These properties were particularly undesirable in stopped-flow experiments since any increase in turbidity in the mixing chamber would be reflected as an increase in absorbance. No turbidity was observed, however, in enzyme solutions upon the addition of lysolecithin.
Furthermore, reaction mixtures containing lysolecithin were completely free of turbidity for at least 1 hour after pyruvate oxidase was reduced by addition of pyruvate.
For this reason lysolecithin was selected as the lipid activator in the stopped-flow experiments. Fig. 8 records the oscilloscope tracings obtained in stoppedflow experiments measuring the rate of reduction of pyruvate oxidase both in the absence and presence of phospholipid.
The curve in the upper picture represents the rate of reduction of pyruvate oxidase by pyruvate in the absence of phospholipid. The 2 cps signal seen in this tracing is of unknown origin and appeared frequently, but not always, during use of the stoppedflow instrument.
The rate constant for reduction of the enzyme in the absence of lipid activator in three separate experiments varied between 0.33 and 0.46 set+. Table III lists the values for the first order rate constant and the half-life for the reduction of pyruvate oxidase in the absence of phospholipids.
The oscilloscope signal shown in the lower portion of Fig. 8 represents the reduction of the enzyme in the presence of lysolecithin. Pyruvate oxidase, in the presence of lysolecithin, is almost completely reduced during the flow period in the instrument. Thus it is difficult to accurately estimate the rate constant for reduction of the oxidase in the presence of lysolecithin. By expanding the time scale on the oscilloscope to more accurately measure the change in transmittance during the first 200 msec of the reaction, it was estimated that reduction was approximately 85% complete within the flow period. We estimate an upper limit of 20 msec for the half-life of the oxidized enzyme in the presence of the phospholipid, substrate, and cofactors based on the oscilloscope signal shown in Fig. 8 (lower picture). Crystalline Pyruvate Oxidase from E. coli. III Vol. 246,No. 6 This leads to a minimum value for the rate constant of the lipidactivated enzyme of 35 see-I.
This rate constant is undoubtedly on the low side; however, it does clearly demonstrate that reduction of pyruvate oxidase by pyruvate is increased at least lOOfold in the presence of lysolecithin. DISCUSSION The activation of pyruvate oxidase by phospholipids and fatty acids meets the criteria established for allosteric effecters, although the activation is unique in some respects because it represents a hybrid situation between the K and V systems as originally defined by Monod,Wyman,and Changeux (6). In the presence of phospholipids, the K, values for pyruvate and TPP in the pyruvate oxidase reaction are altered and the enzyme can be desensitized with respect to the activator as would be expected in a K type allosteric system; however, the effect of phospholipid on the V msI of the reaction is also representative of a V type system.
In addition, the effect of phospholipids on the kinetic parameters of TPP binding are very unusual. A positive allosteric effector of the K type (6) shifts an enzyme from sigmoidal kinetics in the absence of effector to normal saturation kinetics in the presence of effector.
In the case of TPP and pyruvate oxidase, the opposite is true.
The enzyme shows normal saturation kinetics in the absence of activator and shifts to sigmoidal kinetics in the presence of phospholipid. The large stimulation of the V,,, of the pyruvate oxidase reaction by phospholipids is unusual.
In some V type systems, V,,, values in the presence and absence of effecters are not nearly so drastically changed (5-7) ; in other V type systems, the activating effect of the allosteric effector is due to its antagonistic action with respect to allosteric inhibitors (8). In the case of pyruvate oxidase, depending on the assay system used, V,,, increases as much as loo-fold in the presence of the phospholipid effector. The rapid kinetics experiments clearly indicate that the ratelimiting step which is altered by the presence of phospholipids occurs at or before the reduction of enzyme-bound flavin.
Perhaps the simplest interpretation would be that phospholipids promote interaction between TPP and FAD prosthetic groups in the enzyme, and thus increase the over-all Vmax for reduction. (2). Taken altogether, these data suggest that pyruvate oxidase is a tetramer consisting of four identical subunits, each capable of binding both a TPP and FAD molecule and on the average, in the presence of phospholipid, two of the TPP sites would have to be occupied in order to observe significant enzymic activity.
In addition to conferring cooperativity to the enzyme with respect to binding TPP, the phosphatides also lower the K, for TPP 3-to 4-fold and the K, for pyruvate is lowered la-fold upon addition of phospholipids to the reaction mixture. In contrast to TPP, however, the saturation kinetics with respect to pyruvate is normal both in the presence and absence of phospholipids.
Thus, while there undoubtedly is more than one binding site for pyruvate in the tetramer, all sites behave completely independent from each other. The physiological significance of the pyruvate oxidase reaction in the over-all metabolism of E. coli remains obscure at this time.
As diagrammed in Fig. 9, E. coli has three separate paths for production of acetate from pyruvate. The principal path for pyruvate oxidation is the dehydrogenase system since acetaterequiring mutants of E. coli can be readily isolated which are blocked either in the decarboxylase (11) or the transacetylase (12) component of the pyruvate dehydrogenase complex. Since these mutants contain pyruvate oxidase but still require acetate for growth, it follows that pyruvate oxidase is either not present in sufficient quantities to supply the acetate requirement for cellular growth or if present in sufficient quantities, it must be subjected to strict metabolic control and be unable to function effectively under the imposed growth conditions. Dietrich and Henning (13) favor the low level of enzyme hypothesis, whereas results in our laboratory would favor the control hypothesis.