Purification and Properties of the a-Ketoglutarate Dehydrogenase Complex of Cauliflower Mitochondria

SUMMARY A procedure is described by which the oc-ketoglutarate dehydrogenase-lipoyl transsuccinylase portion of the oc-keto-glutarate dehydrogenase complex of cauliflower mitochondria has been pursed to a specific activity >3. The enzymatic activity is totally dependent on ac-ketoglutarate, DPN, reduced CoA, thiamine pyrophosphate, Mg2+ or Ca2+, and lipoyl dehydrogenase. Either cauliflower or pig heart lipoyl dehydrogenase will couple with the cr-ketoglutarate dehy-drogenase-lipoyl transsuccinylase complex to produce DPNH and succinyl-Cob. Because the cauliflower enzyme a requirement for both thiamine-PPi it rather free half-life The for the the lag

A procedure is described by which the oc-ketoglutarate dehydrogenase-lipoyl transsuccinylase portion of the oc-ketoglutarate dehydrogenase complex of cauliflower mitochondria has been pursed to a specific activity >3. The enzymatic activity is totally dependent on ac-ketoglutarate, DPN, reduced CoA, thiamine pyrophosphate, Mg2+ or Ca2+, and lipoyl dehydrogenase.
Either cauliflower or pig heart lipoyl dehydrogenase will couple with the cr-ketoglutarate dehydrogenase-lipoyl transsuccinylase complex to produce DPNH and succinyl-Cob.
Because the cauliflower enzyme as prepared has a requirement for both thiamine-PPi and Mg2+, it was possible to show that the enzyme binds magnesium-thiamine-PPi rather than free Mg2+ and free thiamine-PPi.
The enzyme-magnesium-thiamine-PPi complex is relatively stable, with a half-life of about 2 min.
The K, for magnesium-thiamine-PPi determined on the basis of initial rates was 8.5 X 1OV M, while the K, determined by measurement of the length of the lag between the addition of magnesium-thiamine-PPi and the initiation of the reaction was 6.1 X 10e6 M. This close agreement supports the conclusion that the lag phase found when the reaction is initiated with magnesium-thiamine-PP; is due to the slow formation of the enzyme-magnesium-thiamine-PPi complex. As usually isolated, the oc-ketoglutarate dehydrogenaselipoyl transsuccinylase complex activity is apparently limited by an inadequate level of lipoyl dehydrogenase still complexed with the other two enzymes.
When the ol-ketoglutarate dehydrogenase-lipoyl transsuccinylase was isolated completely free of lipoyl dehydrogenase, it was possible to saturate this complex by the addition of lipoyl dehydrogenase either from cauliflower or from pig heart. When the lipoyl dehydrogenase is present in excess it is possible accurately to determine the kinetic parameters for the substrates and cofactors of the complex.
The oxidation of cr-ketoglutarate is catalyzed by cu-ketoglutarate dehydrogenase ( which has been isolated as a complex with a molecular weight of about 2 x 106, from Escherichia co& pig heart muscle, pigeon breast muscle, and bovine kidney mitochondria (1). This enzyme complex has not previously been obtained in an active soluble form from higher plant tissues.
All of the cr-ketoglutarate dehydrogenase complexes previously isolated have been shown to contain thiamine pyrophosphate, flavin adenine dinucleotide, covalently bound lipoic acid, and a divalent metal ion, all of which are required for the oxidation of or-betoglutarate by diphosphopyridine nucleotide, resulting in the formation of succinyl-CoA as shown in the reaction a-Ketoglutarate + DPN+ + CoA-SH + succinyl-Coil + COZ + DPNH + H+ This reaction proceeds successively via decarboxylation of CXketoglutarate, reductive succinylation of a protein-bound lipoyl moiety, succinyl transfer, and electron transfer reactions as depicted in the sequence of reactions (2) at top of next page.
The complexes which catalyze this sequence of reactions in bacteria and mammals have been resolved into three separate enzyme fractions which can be reconstituted to give active complexes (1 All leaves were removed, and the heads were washed with cold tap water and chilled to lo prior to mitochondrial preparation. The heads were grated with a conical stainless steel grater driven by a high torque motor, arranged so that cell disruption occurred under the surface of the isolation medium.
All steps of mitochondrial and enzyme preparation were carried out between 0 and 4". A crude suspension of mitochondria was obtained by grating 1500 g of cauliflower tissue into 1500 ml of 0.6 M sucrose, 0.001 M MgCL, 0.005 M EDTA, and 0.05 M Tris at pH 7.4. This suspension was strained through a tightly woven nylon bag with the aid of a Pexton press and centrifuged at 10,000 X g for 30 min.
The lighter portion of the sediment was resuspended by gentle swirling in 206 ml of 0.4 M sucrose, 1 X lop4 M EDTA, 0.01 M MgC&, and 0.05 M Tris at pH 7.0. This resuspended material was then centrifuged at 10,000 X g for 30 min. The sedimented material from this centrifugation was resuspended in 40 ml of 1 X 1O-4 M EDTA, 0.01 M MgC12, and 0.05 M Tris, pH 7.0. A second preparation of the same size was routinely carried out in parallel with the first and the combination of the two preparations produced 80 ml of mitochondrial suspension containing about 8.5 mg of protein per ml. Further preparation of the enzyme from this mitochondrial suspension is described under "Results." Assay Procedures-Except as otherwise noted, the activity of the KG-dehydrogenase-lipolyltranssuccinylase complex was assayed by adding 0.005 to 0.01 unit of the complex to a reaction mixture already containing 10 pmoles of Tris-Cl (pH 6.9), 10 pmoles of MgCh, 5 pmoles of Tris DPN, 10 pmoles of dithiothreitol, 0.225 pmole of thiamine-PPi, and 50 units of pig heart LipDH.
This mixture was incubated for 90 set in the cell compartment of the spectrophotometer (controlled at 25")) after which the reaction was started by adding 5 pmoles of Tris-cuketoglutarate and 0.1 pmole of CoA-SH in 25 ~1 of 0.01 M dithiothreitol.
The final volume was 1.0 ml and the final pH was 6.9. The appearance of DPNH was followed at 340 nm. Since the reaction of this enzyme produces 1 mole of DPNH for each mole of a-ketoglutarate oxidized, 1 unit of KG-dehydrogenaselipoyl transsuccinylase is defined as that amount of enzyme which catalyzes the oxidation of 1 Kmole of oc-ketoglutarate per min. The activity of lipoyl dehydrogenase was determined in an assay mixture containing 50 pmoles of Tris-Cl, 2 pmoles of lipoamide, and 0.1 pmole of DPNH in a final volume of 1.0 ml at pH 7.0. The reaction was started with enzyme. The decrease in DPNH absorbance was followed at 340 nm and the rate is expressed as micromoles of DPNH consumed per min. One unit of lipoyl dehydrogenase is defined as that amount of enzyme which catalyzes the oxidation of 1 pmole of DPNH per min.
Protein concentrations were determined by the method (5) of Lowry et al. with crystalline bovine serum albumin as a standard or by calculation from the absorbance of the solution at 280 nm and 260 nm with the method of Warburg and Christian (5). Specific activity of enzymes is expressed as units per mg of protein.
Thiamine pyrophosphate was determined by conversion to thiochrome pyrophosphate by treatment with alkaline ferricyanide.
Determinations were made in a volume of 5.0 ml containing 1.0 ml of 1 N NaOH and 0.1 ml of 0.01 M potassium ferricyanide.
Fluorescence of the thiochrome was read in a Turner model 111 fluorometer with a primary filter with a peak transmission at 360 nm and secondary filters with a peak transmission at 436 nm. The fluorescence of known solutions of thiamine-PPi was found to be linear with concentration. Values of maximal velocity and K, were determined from plots according to the method of Lineweaver and Burk fitted by the procedure of Wilkinson (7) to provide appropriate weights. Plotted lines are those described by the equations generated by that fitting process.

Commercial
Enzymes-Analytical reagent grade lipoamide dehydrogenase (NADHz:lipoamide oxidoreductase, EC 1.6.4.3) from pig heart with a specific activity of 210 i.u. per mg was obtained from Boehringer Mannheim and dialyzed at 4" against three changes of 0.005 M Tris (pH 7.0) containing 1 X 10h5 M EDTA.
When assayed with lipoamide, the resulting pig heart lipoyl dehydrogenase solution contained 1000 units of activity per ml at 25".
Highly purified bovine pancreatic ribonuclease A (type XA, solution in 0.2 M phosphate buffer, pH 6.4) and crystalline bovine pancreatic deoxyribonuclease I (type DN-C) were obtained from Sigma.
Dithiothreitol and crystalline bovine serum albumin were obtained from Calbiochem. The cr-ketoglutaric acid, thiamine pyrophosphate, and oxidized pyridine nucleotide solutions were made to pH 7.0 with Tris. Enzyme grade Tris and enzyme grade (NHSZSOq were supplied by Mann.
All other chemicals used were reagent grade or better.
All solutions were made in double distilled water, deionized before final glass distillation.

RESULTS
Early attempts to solubiliae the a-ketoglutarate dehydrogenase from cauliflower were hampered by a high level of DPNH oxidase in mitochondria and preparations made from mitochondria. It was found that this activity was sensitive to repeated freezing and thawing cycles, and the use of this technique to eliminate DPNH oxidase made a reliable assay for cr-ketoglutarate dehydrogenase possible.
In preliminary efforts at purification of the ac-ketoglutarate dehydrogenase complex it was found that the lipoyl dehydrogenase of the plant complex separates into two fractions during ammonium sulfate fractionation, one separating with the KGdehydrogenase-lipoyl transsuccinylase activity, and the remainder at a much higher ammonium sulfate saturation (see Fig. 1). Attempts to purify the entire ar-ketoglutarate dehydrogenase complex by a procedure involving ultracentrifugation, pH precipitation, protamine precipitation according to Hirashima, Hayakawa, and Koike (8) Rate of cu-ketoglutarate dehydrogenase reaction (reported as micromoles of a-ketoglutarate oxidized per min) as a function of the addition of lipoyl dehydrogenase from cauliflower or pig heart to a-ketoglutarate dehydrogenase-lipoyl transsuccinylase from cauliflower mitochondria.
The assay mixture contained 1 mM DPN, 10 mm phosphate, 0.2 mM thiamine-PPi, 10 mM MgCl,, 10 mg of bovine serum albumin, 2.5 mM cu-ketoglutarate, and 0.1 mM CoA-SH in a total volume of 1 ml. The pH was adjusted to 6.9 with KOH. Each assay contained 2 ~1 of the enzyme from Step II of the purification procedure which included 0.001 unit of lipoyl dehydrogenase activity.
The units of lipoyl dehydrogenase shown include both the additions of purified lipoyl dehydrogenase and that added with the enzyme preparation.
A, cauliflower lipoyl dehydrogenase; 0, pig heart lipoyl dehydrogenase. In preparations such as the 30 to 40% saturation ammonium sulfate precipitate in Fig. 1, in which some lipoyl dehydrogenase activity separated with the KG-dehydrogenase-lipoyl transsuccinylase complex, the activity of the preparation as a function of enzyme concentration was highly nonlinear. This nonlinearity could be abolished by the addition of either cauliflower or pig heart lipoyl dehydrogenase to the assay mixture. This indication of a deficient amount of lipoyl dehydrogenase in preparations which could be designated as "complete" a-ketoglutarate dehydrogenase complex was confirmed by experiments such as that illustrated in Fig. 2, in which a constant amount of KG-dehydrogenase-lipoyl transsuccinylase complex essentially free of lipoyl dehydrogenase was assayed with increasing amounts of lipoyl dehydrogenase from either cauliflower or pig heart. The amount of lipoyl dehydrogenase required to produce a maximum a-ketoglutarate dehydrogenase activity was much greater wit.h the pig heart than with the cauliflower lipoyl dehydrogenase, perhaps indicating a preferential binding of the plant enzyme, but the amount of lipoyl dehydrogenase required was considerably in excess of the levels of that activity which could be isolated in combination with the KG-dehydrogenaselipoyl transsuccinylase complex by any technique tried. Because of difficulties in the purification of the cauliflower lipoyl dehydrogenase, and the inhibition of the over-all reaction produced by impurities in this preparation, pig heart lipoyl dehydrogenase was routinely used for assay of the KG-dehydrogenase-lipoyl transsuccinylase complex in the work reported here.
Purification of a-Ketoglutarate Dehydrogenase-Lipoyl Transsuccinylase Complex The purification scheme which produced enzyme with maximum specific activity and which was used for preparation of the KG-dehydrogenase-lipoyl transsuccinylase complex whose characteristics are reported here is as follows.
Step I: Extraction of Mitochonclria--A mitochondrial suspension prepared as described under "Methods" was made 0.01 M in dithiothreitol and 40-ml aliquots were sonically disrupted in a The supernatant fractions were frozen at -15" for at least 4 hours and then thawed slowly at room temperature.
The freezing-thawing cycle was repeated twice over a period of at least 24 hours. After the third thawing, precipitated protein was removed by centrifugation at 20,000 x g for 60 min.
Step II: Ammonium Sulfate Fractionation-To the suspension from Step I a solution of saturated ammonium sulfate in 0.001 M glycylglycine buffer, pH 7.0, was added dropwise with stirring. The concentration was first brought to 30% saturation (calculated as per cent of the saturated ammonium sulfate in the total volume), stirred for 1 hour, and centrifuged at 20,000 x g for 30 min. The supernatant liquid was brought to 45% saturation by addition of saturated ammonium sulfate, permitted to stir for 1 hour, and centrifuged as before.
The precipitate was resuspended in a minimal volume (approximately 10 ml) of 0.05 M Tris (pH 7.0) containing 1 X lop4 M EDTA.
It was then frozen overnight at -15" and, after thawing, was centrifuged at 20,000 x g for 30 mm to remove the precipitated protein.
Step III: Nuclease Treatment-Ribonuclease and deoxyribonuclease were added to the supernatant solution from Step II to give a final concentration of 25 pg of each nuclease per ml. The mixture was incubated at 25" for 60 min, after which it was placed on a Sephadex G-200 column (2.5 x 40 cm) and eluted with 0.05 M Tris buffer, pH 7.0, containing 1 X 10m4 M EDTA.
Four of the 3.3-ml fractions containing the ac-ketoglutarate dehydrogenase-lipoyl transsuccinylase activity were pooled and centrifuged for 20 mm at 20,000 x g to remove a white colloidal material which forms in the column and elutes with the activity.
Passage through the column separates the KG-dehydrogenase transsuccinylase complex from the nucleases, the nucleotides which they have produced from nucleic acid contaminants, and some other small proteins. B had no effect on the rate (solid line).
Step IV: Second Ammonium Sulfate Fractionation-The supernatant solution from Step III was made to 33% ammonium sulfate saturation with saturated ammonium sulfate solution as before. After stirring for 1 hour, the suspension was centrifuged at 20,000 X g for 30 min and the precipitate was discarded. The saturation with ammonium sulfate was then brought to 40% and the suspension was stirred for 1 hour and allowed to stand overnight. The precipitate was collected by centrifugation and resuspended in a minimal volume (1 to 5 ml) of 0.05 M Tris with 1 X UY4 M EDTA, pH 7.0.
A typical preparation by the steps outlined above is summarized in Table I. The enzyme prepared by this method was free of lipoyl dehydrogenase and DPNH oxidase activity, but contained 0.22 unit of pyruvate dehydrogenase per unit of crketoglutarate dehydrogenase activity at the time of isolation. Attempts at further purification with sucrose density gradient techniques did achieve some separation of the a-ketoglutarate dehydrogenase and pyruvate dehydrogenase activities with a further reduction of the pyruvate dehydrogenase contaminant, but the separation was not complete. The pyruvate dehydrogenase activity disappears during storage, and is not detectable after about 3 weeks of storage.
The KG-dehydrogenase-lipoyl transsuccinylase complex prepared in the manner described above is quite stable and can be stored at -15" for several months without significant loss of activity. Preparations stored for more than 1 year retained 80% of the original activity. The enzyme appears to be unaffected by freezing and thawing and loses no activity when stored at room temperature for several hours. When the preparation is assayed without the addition of dithiothreitol, there is an apparent loss of activity with time, but this is completely reversed by the addition of dithiothreitol to the assay medium. When assayed in the presence of saturating levels of pig heart lipoyl dehydrogenase, the activity is linear with respect to concentration of the KG-dehydrogenase-lipoyl transsuccinylase complex. The purif?.ed complex of cr-ketoglutarate dehydrogenase and lipoyl transsuccinylase isolated from cauliiower mitochondria is in many respects similar to the enzymes previously isolated from bacterial and mammalian sources. It is totally dependent on or-ketoglutarate, CoA-SH, thiamine-PPi, DPN, and lipoyl dehydrogenase. There were no detectable rates in the absence of any of these factors and starting the reaction with any of them resulted in identical rates. The course of a typical assay is shown in Fig. 3. The addition of either DPN, a-ketoglutarate, CoA-SH, or lipoyl dehydrogenase at Point A after 90.set preliminary incubation of the enzyme in the presence of the other reaction components resulted in the initiation of the reaction, with the establishment of a rate which was linear for about 1 min. The rate of DPNH production then decreased to a slower second rate which was constant for 10 min or longer. The addition of a second aliquot of a-ketoglutarate, DPN, or lipoyl dehydrogenase at Point B had no effect on the rate then established. However, a second addition of an equal amount of CoA-SH resulted in an increased rate as indicated by the dashed line of Fig. 3.
The initial rate resulting from the second CoA-SH addition was always slower than that found when the reaction was first initiated and the total DPNH produced before the establishment of the second rate was less than after the first addition of CoA-SH. These observations suggest product inhibition by succinyl-CoA, and this compound has been found to inhibit the initial rate of the reaction. The addition of either Mg* or Ca* was found to stimulate the activity of the cauliflower KG-dehydrogenase-lipoyl transsuccinylase complex in our assays, as has been reported for both bacterial and mammalian enzymes (3). Determination of calcium and magnesium in our preparations by atomic absorption revealed that the assay mixtures contained approximately 1 x lop5 M each of magnesium and calcium before the addition of any metal ions. Most of the ions present in the assay mixture were contributed by the Tris salt and dithiothreitol.
In view of this significant contaminant of metal ions it was difficult to be certain that there was an absolute requirement for M@+ or CaZt-. This requirement was shown by the addition of 3.3 x 10h5 M EDTA to the reaction mixture with the results shown in Table  II. It may be seen that EDTA essentially abolishes the ac- Bound lipoate was not determined, but the specificity of pig heart lipoyl dehydrogenase for lipoate is strong evidence for the presence of bound lipoate in the cauliflower complex.
The oc-ketoglutarate dehydrogenase-lipoyl tranesuccinylase complex was completely precipitated by centrifugation at 144,000 X g for 2.5 hours., This compares with the 2 hours required to sediment the entire oc-ketoglutarate dehydrogenase complex isolated from pig heart by Sanadi (9) which has a molecular weight of 2 x 106. In addition, the cauliflower KGdehydrogenase-lipoyl transsuccinylase complex was eluted in the void volume of Sephadex G-200, which excludes all proteins with a molecular weight over 800,000 (10). Both of these observations suggest a high molecular weight, although no direct molecular weight determinations have been made. The effect of pH on cr-ketoglutarate dehydrogenase-lipoyl transsuccinylase activity was determined in a mixed buffer system containing both 0.01 M Tris-Cl and 0.01 M potassium phosphate.
Determinations at the pH extremes were repeated with additional lipoyl dehydrogenase to ensure that the rate was not limited by an effect of pH on the lipoyl dehydrogenase. The mixed buffer gave the same rate as either buffer alone at the same pH. The effect of pH on the reaction rate is shown in Fig. 4, where the pH indicated for each point is that measured after completion of the assay. The pH optimum in 6.9, which is somewhat lower than that reported for the bacterial enzyme, which it otherwise resembles more closely than that from pig heart (see Table III).

Interaction of Metal Ions, Thiamine-PPi, and Enzyme Complex
Preliminary studies with nonsaturating thiamine-PPi concentrations showed a distinct lag phase when the reaction was initiated by thiamine-PPi, Mgz+, or enzyme which was not found when the enzyme was previously incubated with both thiamine-PPi and Mg2+. This lag phase is illustrated in Curve -4 of Fig. 5. If the enzyme is first incubated for 90 set with thiamine-PPi and Mg"+, the reaction starts immediately on the addition of the previously incubated enzyme (Line B of Fig. 5). If the previously incubated enzyme is added to an assay mixture which contains no thiamine-PPi and Mg* other than that added with the enzyme, the reaction starts at the same time, but slows more rapidly than when thiamine-PPi and metal are provided (Line C of Fig. 5). This observation suggests the formation of a relatively stable enzyme-metal-thiamine-PPi complex. The stability of this complex was studied by incubating the KG-dehydrogenase-lipoyl transsuccinylase complex for 24 hours at 0" in 0.05 M Tris (pH 7.0) containing 0.01 M Mg2+ and 1 x lo+ M thiamine-PPi.
Aliquots of 5 ~1 of this previously incubated mixture were then added to l-ml assay mixtures lacking thiamine-PPi and incubated at 25" for periods up to 10 min. After various intervals of incubation the reaction was started by adding a mixture of CoA-SH and ol-ketoglutarate as for the standard assay. A similar series of assays were previously incubated at 25" in the absence of both Mg2f and thiamine-PPi.
If a relatively stable complex were formed during the preliminary incubation at O", the decrease in rate found when it was subsequently diluted and incubated without additional Mg2f or thiamine-PPi would reflect dissociation of that complex.
As is shown in Fig. 6, when the natural logarithm of the initial rate of the reaction is plotted against the incubation time at 25", the indicated pseudo-first order rate constant for decay of the complex (obtained from the slope of lines fitted by the method of least squares) is 0.0094 set-1 in the absence of Mg2+ and 0.0025 set-1 when Mg2+ is present during the 25" The delay in the initiation of the oxidation of ar-ketoglutarate by the cauliflower complex which is found if the enzyme is not previously incubated with thiamine-PPi and Mg"f offers an opportunity to evaluate the second order rate constant for the formation of the metal-thiamine-PPi-enzyme complex. Dixon and Webb (15) have derived equations for the relationship between the second order rate constant for the formation of an enzyme complex and the time elapsed before the establishment of the steady state for its reaction.
They have defined T, the length of the lag phase, as the time from initiation of the reaction to the point at which extrapolation of the linear steady state rate intersects the extrapolated zero product concentration line (see Fig. 5). In reciprocal form the equation of Dixon and Webb is where X0 is the initial concentration of the compound which binds to the enzyme.
In this case, because of evidence presented below, this is considered to be magnesium-thiamine-PPi, the formation of which from Mg* and thiamine-PPi is assumed to be much more rapid than its joining in complex with the enzyme.
In the data presented in Fig. 7, the length of the lag phase, 7, was measured by starting the standard assay with thiamine-PPi after a 90-set preliminary incubation of the enzyme in the otherwise complete reaction mixture.
A similar series of assays was run with 0.01 M CaCh substituted for 0.01 M MgC12. As predicted by the equation above, the results with either ion give straight lines when l/r is plotted against thiamine-PPi concentration.
On the basis of the dissociation constant determined for magnesium-thiamine-PPi (see below), the  (7). thiamine-PPi is presumed to be essentially completely in the form in complex under the conditions of these assays, and the same is probably true for the calcium-thiamine-PPi as well. With this assumption, the second order rate constants for the formation of enzyme-magnesium-thiamine-PPi and enzymecalcium-thiamine-PPi complexes were found to be 687 10 set-l and 4469 M-~ set-I, respectively, as indicated by the slopes of the fitted lines of Fig. 7. According to the equation, the slope of these lines is equal to l~+~, the rate constant for the formation of enzyme-metal-thiamine-PPi complex. These data also provide an opportunity for an independent evaluation of the K,,, for the magnesium-thiamine-PPi substrate of the reaction, which according to Dixon and Webb's equation is equal to the intercept divided by the slope of the l/r against magnesium-thiamine-PPi lines. The values obtained in this way for the K,,, values were 6.1 X 10e6 M for magnesium-thiamine-PPi and 4.0 X 10e6 M for calcium-thiamine-PPi.
When Km values are calculated from the linear rates measured in the same experiments by fitting weighted limes to the reciprocal of the rate against the reciprocal of the magnesium-thiamine-PPi concentration in a conventional double reciprocal plot, the values were found to be 8.6 x lop6 M for magnesium-thiamine-PPi and 1.0 x 10v6 M for calcium-thiamine-PPi.
When the Km values derived in the two different ways are compared, this agreement lends support to the conclusion that the lag phase is the result of the slow formation of a complex of metal ion and thiamine-PPi with the enzyme.
Because the cauliflower KG-dehydrogenase-lipoyl transsuccinylase complex is isolated without bound thiamine-PPi or magnesium, it has been possible to investigate the question of whether Mg* and thiamine-PPi bind independently to the enzyme either randomly or in an ordered sequence or whether a magnesium-thiamine-PPi complex is first formed which then binds to the enzyme.
For this purpose the apparent dissociation constant of magnesium-thiamine-PPi was determined under the conditions used for the enzyme assay. This constant, Kd = 4.07 x 10m4 M, is of the same order of magnitude as has been determined for magnesium-ADP (16).
This Kd has been used to calculate the apparent magnesiumthiamine-PPi concentration at different levels of Mg* and thiamine-PPi.
The cr-ketoglutarate oxidation rates found in the presence of Mg2+ concentrations from 10h4 to lop2 M in combination with thiamine-PPi concentrations ranging from 3.0 X low6 M to 3.75 X low3 M were plotted aginst the calculated magnesium-thiamine-PPi concentration for each combination. In these experiments the standard assay procedure, including a 90-set preliminary incubation of enzyme with magnesium and thiamine-PPi, was used, and the reactions started by adding a-ketoglutarate and CoA-SH. Within the range indicated, the rates all fall essentially on the same line when l/v is plotted against l/magnesium-thiamine-PPi concentration as shown in Fig. 8. This strongly suggests that magnesium-thiamine-PPi is in fact the form which binds to the enzyme during the preliminary incubation, and the absence of significant inhibition when either free Mg* or free thiamine-PPi is high relative to the concentration of magnesium-thiamine-PPi indicates that the free forms do not bind significantly to the enzyme.
In this experiment the apparent K, for magnesium-thiamine-PPi was found to be 1.1 X 10v5 M.

DISCUSSION
The cY-ketoglutarate dehydrogenase complex of cauliflower mitochondria has proven to be impossible to isolate in its entirety by the methods attempted.
Any complete complex which was purified contained levels of lipoyl dehydrogenase inadequat,e to saturate fully the potential activity of the preceding enzymes in the complex.
This may be true of the complex isolated from other organisms, and may even represent the situation in viva. The variability in the amount of lipoyl dehydrogenase which comes through various steps of purification still attached to the remainder of the complex, and the very high levels of lipoyl dehydrogenase activity which are found separately render any conclusions on this point uncertain.
We have chosen to purify only the ol-ketoglutarate dehydrogenase-lipoyl transsuccinylase portion of the complex and to assay this under uniformly saturating levels of lipoyl dehydrogenase provided separately. We believe that these conditions provide a more reliable means for studying the activity of the KG-dehydrogenase-lipoyl transsuccinylase complex and accordingly give more reliable estimates of the properties of this complex.
The higher plant complex isolated in this way is obtained with a specific activity which approaches that of the enzyme previously purified from mammalian and bacterial sources. As with those complexes (16), there is some contamination of pyruvate dehydrogenase.
However, since these activities can be separated to some degree by sucrose gradient centrifugation of the purified enzyme from Step IV of the procedure, and since the ar-ketoglutarate dehydrogenase activity is stable while the pyruvate dehydrogenase rapidly disappears in storage, there is little question that the pyruvate dehydrogenase activity is due to a different enzyme and is not a minor activity of the (Yketoglutarate dehydrogenase complex. The relative ease with which lipoyl dehydrogenase dissociates from the remainder of the KG-dehydrogenase complex and the high concentration of apparently unbound lipoyl dehydrogenase may indicate that the same situation obtains in the intact cell, with lipoyl dehydrogenase in loose association with the various cr-keto acid dehydrogenases, perhaps without specificity as to the preferred complex (12). However, the higher level of activity obtained from the addition of cauliflower lipoyl dehydrogenase to the cauliflower KG-dehydrogenase-lipoyl transsuccinylase complex, as compared with t.hat provided by the pig heart lipoyl dehydrogenase, indicates that the plant lipoyl dehydrogenase either binds better to the plant complex or for other reasons forms a better coupling system with it. The nature of the binding of the lipoyl transsuccinylase to lipoyl dehydrogenase is presently unclear. It may be that only the reduced lipoate of lipoyl transsuccinylase must bind to the lipoyl dehydrogenase, or it may be, as Reed (17) suggesk, that the lipoyl dehydrogenase must itself be bound to the lipoyl transsuccinylase before it can accept the lipoate of that enzyme. One can speculate that either mechanism is possible, and that the reason for the lower activity with the pig heart LipDH is that it does not fit the binding site for lipoyl dehydrogenase on the lipoyl transsuccinylase and must accept the lipoate from an unfavorable configuration.
The absence of bound thiamine-PPi in the cauliflower complex has made it possible to show that the cofactor for this enzyme appears to be magnesium-thiamine-PPi rather than magnesium and thiamine-PPi separately.
The magnesium-thiamine-PPi dissociates readily, although rather slowly, from the enzyme. This dissociation of a required cofactor may represent one method of controlling the activity of KG-dehydrogenase in plants.