Cofactor Requirements of γ-Butyrobetaine Hydroxylase from Rat Liver

The hydroxylation of y-butyrobetaine (4-trimethylaminobutyrate) to carnitine (3-hydroxy-4-trimethylaminobutyrate) is catalyzed by a soluble enzyme from rat liver which has been partially purified. The enzyme which previously has been shown to require molecular oxygen and ferrous ion has a specific requirement for 2-ketoglutarate. Several reductams stimulate the formation of hydroxylated product; the most active ones are ascorbate and isoascorbate, whereas reduced 2,6-dichlorophenolindophenol and 2-amino-5,6dimethyl-4-hydroxy-5,6,7,%tetrahydropteridine are less active. The previously observed stimulation both by NADPH, isocitrate, isocitrate dehydrogenase (EC 1.1.1.42), and by microsomes has been found to be related to the formation of 2-ketoglutarate. One mole of 2-ketoglutarate is degraded per mole of carnitine formed. Carbon dioxide and succinate are products of 2-ketoglutarate degradation; no free succinic semialdehyde can be detected. Several compounds which are chemically or biologically related to y-butyrobetaine and to 2-ketoglutarate have been tested as inhibitors in the reaction. Only succinic semialdehyde and 3-trimethylaminopropyl-I-sulfonate are effective inhibitors. The sulfhydryl reagents fi-chloromercuribenzoate and p-chloromercuriphenylsulfonate in 0.1 mM concentration completely inhibit the formation of carnitine after preliminary incubation with the enzyme for 20 min at 37”. N-Ethylmaleimide, iodosobenzoate, and iodoacetate are less effective inhibitors under these conditions, whereas sodium arsenite, carbarsone, and acetarsone cause appreciable inhibition only in 10 mu concentration. Preliminary incubation of the enzyme with ferrous ion and ascorbate results in signiticantly lower yield of carnitine than when catalase has also been included in the preliminary incubation medium. It is suggested that y-butyrobetaine is hydroxylated to carriitme simuitaneously wXn the oxitiative aecarboxylation of P-ketoglutarate in a reaction sequence which involves the intermediate formation of a peroxide of the two substrates. Ferrous ion might act as oxygen-activating agent. Free sulfhydryl groups are apparently necessary for enzymic activity and ascorbate and catalase probably act by maintaining these groups as well as ferrous ion in the reduced state.

to carnitine (3-hydroxy-4-trimethylaminobutyrate) is catalyzed by a soluble enzyme from rat liver which has been partially purified.
The enzyme which previously has been shown to require molecular oxygen and ferrous ion has a specific requirement for 2-ketoglutarate. Several reductams stimulate the formation of hydroxylated product; the most active ones are ascorbate and isoascorbate, whereas reduced 2,6-dichlorophenolindophenol and 2-amino-5,6dimethyl-4-hydroxy-5,6,7,%tetrahydropteridine are less active. The previously observed stimulation both by NADPH, isocitrate, isocitrate dehydrogenase (EC 1.1.1.42), and by microsomes has been found to be related to the formation of 2-ketoglutarate.
One mole of 2-ketoglutarate is degraded per mole of carnitine formed.
Carbon dioxide and succinate are products of 2-ketoglutarate degradation; no free succinic semialdehyde can be detected.
Several compounds which are chemically or biologically related to y-butyrobetaine and to 2-ketoglutarate have been tested as inhibitors in the reaction. Only succinic semialdehyde and 3-trimethylaminopropyl-I-sulfonate are effective inhibitors.
The sulfhydryl reagents fi-chloromercuribenzoate and p-chloromercuriphenylsulfonate in 0.1 mM concentration completely inhibit the formation of carnitine after preliminary incubation with the enzyme for 20 min at 37". N-Ethylmaleimide, iodosobenzoate, and iodoacetate are less effective inhibitors under these conditions, whereas sodium arsenite, carbarsone, and acetarsone cause appreciable inhibition only in 10 mu concentration. Preliminary incubation of the enzyme with ferrous ion and ascorbate results in signiticantly lower yield of carnitine than when catalase has also been included in the preliminary incubation medium. It is suggested that y-butyrobetaine is hydroxylated to carriitme simuitaneously wXn the oxitiative aecarboxylation of P-ketoglutarate in a reaction sequence which involves the intermediate formation of a peroxide of the two substrates. Ferrous ion might act as oxygen-activating agent. Free sulfhydryl groups are apparently necessary for enzymic activity and ascorbate and catalase probably act by maintaining these groups as well as ferrous ion in the reduced state. y-Butyrobetaine (4-trimethylaminobutyrate) is hydroxylated to carnitine (3-hydroxy-4-trimethylaminobutyrate) by the rat and the mouse (3-5), and a soluble protein fraction catalyzing this reaction has been obtained from rat liver homogenates (6). The rate-limiting step is probably the dissociation of the carbonhydrogen bond, as a kinetic hydrogen isotope effect was found with y-butyrobetaine labeled with tritium in the chain (6). An o'xygenase type of reaction is indicated by the requirement for molecular oxygen (6, 7) and evidence has been presented for the requirement for ferrous ion (6,8). A requirement has been observed for a NADPH-regenerating system, vix. isocitrate + isocitrate dehydrogenase (6). Ascorbate, catalase, and microsomes from rat liver and kidney stimulated the formation of hydroxylated product (6). Further studies on the cofactor requirement will now be presented. A solution of 2-ketoglutaric acid was obtained by passing a solution of the zinc salt through a small column of Dowex AG50 W-X8 (hydrogen form) which was then eluted with water. nn-Carnitine chloride, 3-trimethylaminopropionate, 4-dimethylaminobutyrate, 4-trimethylaminocrotonate, 5-dimethylaminovalerate, 5-trimethylaminovalerate, 6-trimethylaminocaproate, and 4-trimethylaminobutan-l-01 were synthesized as described previously (5, 6). Succinic semialdehyde was synthesized according to the method of Carriere (9) as modified by Albers (cited by Jacoby in Reference 10) and was characterized as the 2,4-dinitrophenylhydrazone and by gas chromatography-mass spectrometry of the trimethylsilyl ether of the 0xime.l 3-Dimethylaminopropyl-1-sulfonic acid was prepared by treating 3-aminopropyl-1-sulfonic acid with formaldehyde and formic acid (6, 11). The product was crystallized twice from hot ethanol, m.p. 224,227" with decomposition. CIH~ZNSOZ (167.2) Calculated: H,7.84;N,8.38 Found2 : H,7.85;N,8.28 3-Trimethylaminopropyl-1-sulfonic acid was obtained by treating the corresponding dimethylamine in methanol-water (7 :3) with a lo-fold excess of methyl iodide.
The theoretical amount of 1 M potassium hydroxide was slowly added during 24 hours. The reaction mixture was then taken to dryness, redissolved in water, and passed through a column of AG llA8, using water for the elution.
The fractions containing trimethylaminopropyl-lsulfonate were free from the dimethylamino acid, as judged by thin layer chromatography (see below). 3-Trimethylaminopropyl-1-sulfonic acid chloride was recrystallized from waterethanol, m.p. above 300". 2-Ketoglutaric acid was purified by silicic acid chromatography (see below and legend to Fig. 3). Some preparations contained 30% or more of impurities which were eluted with higher concentrations of tert-butanol than required for 2-ketoglutaric acid. The fractions containing 2-ketoglutaric acid were lyophilized, redissolved in water, and distributed in several ampoules which were then stored at -15". Rechromatography regularly disclosed the presence of up to 5% l To be published.
2 Microanalyses were carried out by The Scandinavian Microanalytical Laboratory, Ballerup, Copenhagen, Denmark. of more polar material which was probably polymerization products.
Silicic acid chromatography did not reveal any impurities in the labeled succinic acid.

Chromatography
Procedures-Ion exchange chromatography on columns of Dowex AG50 W-X8 (minus 400 mesh, hydrogen form) was carried out as described previously (12). Betaines were separated from dimethylamino acids by filtration through a column of AG 1 IA8 (self-adsorbed form) (13). Partition chromatography of carboxylic acids was carried out on columns of silicic acid with 0.25 M sulfuric acid as the stationary phase and benzene with varying concentrations of tert-butanol as the mobile phase according to Prior Ferraz and Relvas (14). For conditions see legend to Fig. 3. Descending paper chromatography of the dinitrophenylhydrazones of 2-ketoglutaric acid and succinic semialdehyde was carried out with 1-butanol-95 y0 aqueous ethanol-O.5 M ammonia in water (13:2:5) as the mobile phase (15). RF values were 0.22 for 2-ketoglutaric dinitrophenylhydrazone and 0.56 for succinic semialdehyde dinitrophenylhydrazone.
Succinic acid was separated from 2-ketoglutaric acid by descending paper chromatography with iso-amyl alcohol saturated with 4 M formic acid as the mobile phase. RF values were 0.44 for 2-ketoglutaric acid and 0.68 for succinic acid. Munktell filter paper No. 312 was used throughout.
Thin layer chromatography was carried out on Silica Gel G with methanoldioxane-25 '% aqueous ammonia (30 : 45 : 25) as the mobile phase, (16). The spots were made visible with iodine vapor.
Approximate RF value for 3-aminopropyl-1-sulfonic acid was 0.33, 0.51 for 3-dimethylaminopropyl-1-sulfonic acid and 0.10 for 3-trimethylaminopropyl-l-sulfonic acid. Rat Liver Homogenates-Male rats of the Sprague-Dawley strain, weighing 200 to 300 g, were used. The animals were killed by a blow on the head and the livers were immediately excised and cooled in ice-cold 0.25 M sucrose. Homogenates (33% wet weight per volume) were prepared in cold 0.25 M sucrose with 0.03 M nicotinamide in a Potter-Elvehjem homogenizer with a tightly fitting Teflon pestle. The tissue was cut into approximately 5-mm pieces and homogenized in the cold at 1,400 rpm for two 45.set periods with an interval of 2 min, during which the tube was cooled in ice. For the preparation of the partially purified enzyme, the supernatant fraction was used after centrifugation at 40,000 x g for 20 min and then at 100,000 X g for 60 min. An MSE type 65 preparative ultracentrifuge was used, and the given g values are average ones. For the preparation of microsomes, the supernatant fraction after centrifugation at 15,000 X g for 20 min was centrifuged for another 60 min at 100,000 X g. The sediment was rehomogenized for 1 min in 0.25 M sucrose with nicotinamide and recentrifuged at 100,000 X g for 60 min. The procedure was repeated another time. The sediment was then rehomogenized in the sucrose solution and the protein concentration was adjusted to approximately 20 mg per ml. The microsomes were prepared fresh for each series of experiments.
Enzyme Preparation-The procedure used previously (6) was slightly modified.
As before, the supernatant fraction of a rat liver homogenate (see above) was fractionated by the addition of a saturated ammonium sulfate solution at 4'. The protein fraction between 40 and 70% saturation was desalted by filtration through a Sephadex G-25 column (coarse) with 20 mM phosphate buffer at pH 6.5 as eluent and then immediately applied to a column of hydroxylapatite in the same buffer (50 g of hydroxylapatite for 2 to 5 g of protein). The column was eluted with a slightly convex concentration gradient of potassium phosphate buffer at pH 6.5. The bulk of the protein emerged from the column between 50 and 150 mM, after which most of the hydroxylase activity was obtained between approximately 175 and 225 mM phosphate buffer (Fig. 1). The hydroxylapatite chromatography resulted in a purification of about six to eight times with about 70% yield.
The partially purified enzyme preparation usually had a specific activity of 3 munits per mg. The enzyme was fairly stable, and could be kept frozen for several months with a decrease in specific activity of less than 50%.
Assay-The assay procedure which has been described in detail previously (6) was used with some modifications.
EFFLUENT VOLUME, ml FIG. 1. Chromatography at 4" on hydroxylapatite (50 g, column dimensions 2.4 X 10 cm) of the 40 to 70% ammonium sulfate fraction of the 100,000 X g supernatant fraction of a rat liver homogenate (2.0 g of protein).
The column was eluted first with 20 mu phosphate buffer at pH 6.5, then with increasing concentration of the buffer (from 20 to 600 m nhosnhate buffer at DH 6.5 Enzymic activity with [methyl-W&-butyrobetaine and [1-r4Crj2-ketoglutarate as substrates was determined as described under "Assay." In the incubations with NADPH, the concentration of NADPH was 0.2 or 1.0 mM.
In the incubations with a NADPH-regenerating system the concentration of NADPH was 0.2 mM. Per ml of incubation mixture there were added also either 0.02 mg of isocitrate dehydrogenase, 4 pmoles of nn-isocitrate, and 1.3 pmoles of MgClz or 5 mg of glucose dehydrogenase and 100 pmoles of glucose or 0.05 mg of glucose 6-phosphate dehydrogenase, 4 pmoles of glucose 6-phosphate, and 6 pmoles of magnesium chloride.
Incubations with a NADPH-regenerating system were carried out both with 20 nnvr phosphate buffer at pH 7.0 and with 20 mM Tris-HCl buffer, pH 7.8. All these incubations contained 40 mM nicotinamide.
In the preliminary incubation experiments given in Table I, 2 mg of glutamate dehydrogenase, 4 pmoles of NADPH, 10 pmoles of ammonium chloride or 0.5 mg of glutamate-oxalacetate transaminase, and 10 pmoles of L-aspartate were used. The commercial suspensions in ammonium sulfate solutions of glutamate dehydrogenaee and glutamate-oxalacetate transaminase had been dialyzed against an excess of 0.1 M phosphate buffer, pH 7.0, at 4' overnight.
In incubations with microsomes, 2 to 5 mg of microsomal protein were added per ml of incubation mixture.
In the studies of the effect of different pH values on the formation of carnitine, the concentration of the phosphate buffer was 50 nnvr. The pH values were recorded at room temperature immediately after the incubations. During the incubations the changes in pH value were less than 0.1 unit.
The incubations were terminated by the addition of an equal volume of 10% trichloraeetic acid. After cooling in ice for 1 hour or overnight the protein was spun down, and the supernatant fractionated on columns of Dowex AG50 W-X8 (minus 400 mesh, H+ form).
The columns were eluted with 0.1 M hydrochloric acid and the amount of radioactivity in the carnitine and y-butyrobetaine fractions were determined with a methaneflow proportional counter. In incubations with succinic semialdehyde and [5-%i]2ketoglutarate, the dinitrophenylhydrazones were prepared as described previously (17). The dinitrophenylhydrazones were separated by means of paper chromatography (see above) and the amount of radioactivity determined with a paper strip counter. The quantitative determination of succinic semialdehyde was carried out according to the method of Bessman,Rossen,and Layne (18) and Prescott and Waelsch (19). In the incubations with [1-l*C&ketoglutarate, the tubes were stoppered, and a 0.5 cm2 filter paper was attached to a piece of wire in the stopper; 1 M solution of Hyamine in methanol (200 ~1) was applied to the filter paper. The incubations were terminated by the addition of either trichloracetic acid or 2 M sulfuric acid, and the diffusion of labeled carbon dioxide was allowed to proceed for 1 hour at 37". The filter papers were then transferred to a scintillation counter vial, containing 16 ml of a mixture of the following composition: 10 g of 2,5-diphenyloxazole, 0.3 g of 1,4-bis[2-(4-methyl-5-phenyloxazolyl)]benzene, 1000 ml of toluene, and 600 ml of methyl cellosolve.
Protein was determined according to the method of Lowry et al. (20) with human serum albumin as standard.
The protein concentration after the hydroxylapatite chromatography was followed by measurement of the absorbance at 280 nm.

RESULTS
Enzyme Preparation- Fig.   1 shows the purification of the hydroxylase by hydroxylapatite chromatography. The y- In Experiment 1, the partially purified protein fraction (12 mg) was incubated for 2 hours at 37" with [methyl-W&butyrobetaine, ferrous ion, ascorbate, catalase, potassium phosphate buffer (pH 7.0), and the additions given in the table. In incubations with 2-ketoglutarate, 10 pmoles were used and the total volume was 4.0 ml. In Experiment 2, the partially purified protein fraction (1.5 mg) and 2-ketoglutarate (0.5 pmole) were incubated in 0.3 ml for 20 min at 37" in phosphate bnffer alone or with the aminating system given in the table. The complete system minus enzyme and 2-ketoglutarate was then added and the incubation (0.8 ml) was continued for 2 hours. See"Assay" for details. There was a linear relationship between the y-butyrobetaine-hydroxylating activity and the concentration of protein in the incubations with the "complete system" of cofactors (see 'Assay").
Requirement for SKetoglutarate as Cofactor-In the previous paper (6) a requirement was reported for a NADPH-regenerating system, viz. NADPH, isocitrate dehydrogenase, and isocitrate. From studies on the specificity of this NADPH-regenerating system it then became apparent that the stimulation was not related to the reduction of NADPH, but to the formation of 2ketoglutarate (Table I) since (a) other NADPH-regenerating systems were inactive, (b) 2-ketoglutarate could replace the isocitrate dehydrogenase system, and (c) very low hydroxylating activity could be observed after preliminary incubation with two different enzymes which catalyze the conversion of Z-ketoglutarate to glutamate.
The hydroxylating activity was also reduced after preliminary incubation of the soluble fraction of a rat liver homogenate under similar conditions before the incubation with y-butyrobetaine and the "complete system" of cofactors (minus 2-ketoglutarate).
In this case glutamate-oxalacetate transaminase plus aspartate caused almost complete inhibition whereas only about 50% inhibition was noted with glutamate dehydrogenase, ammonium chloride, and NADPH, probably because -XL4DPH was reoxidjzed during Ahe incubation.
Several compounds were tested in 0.15 m&c and 1.5 mM concentration for their ability to replace 2-ketoglutarate as a cofactor, tiz. oxalacetate, 2-ketoadipate, 2-ketopimelate, Z-hydroxyglutarate, succinic semialdehyde, glutarate, rm-glutamate, (3 pCi, 0.2 pmole) for 1 hour at 37" with the purified protein fraction (2 mg), r-butyrobetaine, ferrous ion, ascorbate, catalase, and phosphate buffer in 0.8 ml as described under "Assay." Succinic acid (10 mg) wasadded to a 0.2~ml aliquot of the incubation mixture, which then was acidified with 50 pmoles of sulfuric acid and applied to the column, which was eluted with increasing concentrations of bert-butanol in benzene. The eluate was collected in 5-ml fractions. Aliquots of 0.5 ml were dried on glass planchets and the radioactivity was determined with a methane-flow proportional counter. Since none of the tested compounds could replace 2-ketoglutarate there is a highly specific requirement for 2-ketoglutarate as a cofactor in the hydroxylation of y-butyrobetaine to carnitine. An apparent K, value for 2-ketoglutarate was calculated to about 0.5 mM (Fig. 2). Degradation of %Ketoglutarate to Sue&ate- Fig.  3 shows a silicic acid chromatogram of the products obtained after an incubation with [5-zsC&&ketog1utarafe.
One major metabolite was found, which was eluted together with added unlabeled succinic acid. The radioactive material had the same RF value as succinic acid on paper chromatography (see "Experimental Procedures"), and three recrystallizations together with unlabeled The incubations (0.8 ml), which were carried out at 37" for 60 min in stoppered tubes, contained the partially purified protein fraction (8.2 mg), ferrous ion, catalase, phosphate buffer (see "Assay"), different concentrations of [methyl-**C&-butyrobetaine, and [1-X&J2-ketoglutarate. The values are from one experiment in which the 2-ketoglutarate concentration was 6.3 mM and the r-butyrobetaine concentration was 0.38 to 2.5 mM, and another experiment in which the 2-ketoglutarate concentration was 0.63 to 6.3 mM and the r-butyrobetaine concentration was 2.5 mu. Trichloracetic acid (10%) was added to the stoppered tubes after the incubations, and diffusion of 'GO2 onto pieces of filter paper with Hyamine was allowed to proceed for 1 hour at 37". The conversion of r-butyrobetaine to carnitine was then determined by ion exchange chromatography (see "Assay").
succinic acid showed unchanged specific radioactivity of the succinic acid. The other labeled material was eluted at the rate expected for 2-ketoglutarate and its nonenzymic polymerization products (see "Experimental Procedures"). It was ascertained by separate experiments that labeled succinate was not metabolized under the present conditions of incubation, as judged by silicic acid chromatograms.
Similar results have previously been obtained with a y-butyrobetaine hydroxylase from Pseudomonas sp. AK 1 (21). Fig. 4 shows the relationship between the formation of labeled carbon dioxide and the formation of carnitine in incubations with different initial concentrations of y-butyrobetaine and of 2-ketoglutarate.

Formation of Carbon Dioxide in d-Ketoglutarate Degradation-
The data indicate a stoichiometric relationship between the degradation of 2-ketoglutarate and the hydroxylation of y-butyrobetaine.
No enzymic formation of carbon dioxide was found when y-butyrobetaine had been omitted from the incubation mixture.
In a previous study of the bacterial y-butyrobetaine hydroxylase (21), a stoichiometric relationship was also found between the hydroxylation of ybutyrobetaine and the consumption of 2-ketoglutarate. Exclusion of Succinic Semialdehyde as Free Intermediate in Formation of Succinate from d-Ketoglutarate-There was no appreciable oxidation of succinic semialdehyde in a 1.25 mM solution under the incubating conditions used. No labeled succinic semialdehyde was detected in incubations with [5-*4C1]2-ketoglutarate to which had been added unlabeled succinic semialdehyde, i.e., less than 0.3% of the degraded 2-ketoglutarate could be recovered as free succinic semialdehyde.
We have previously reported similar results from experiments with the bacterial y-butyrobetaine hydroxylase (21).
Inhibition by Compounds Structurally Similar to %Ketoglutarate and to y-Butyrobetaine- Table  II shows that slight inhibition was observed when structural analogues of 2-ketoglutarate were added to the incubations. Succinic semialdehyde and 3-trimethylaminopropylsulfonate were effective inhibitors, whereas 3-trimethylaminopropionate and 4-dimethylaminobutyrate were less inhibitory.
The nature of the inhibition has not been investigated.
No inhibition was observed with 4-trimethylaminobutanol.
Ascorbate- Table  III and Fig. 5 show the effect of adding different reductants to incubations with the partially purified protein fraction, 2-ketoglutarate, ferrous ion, and catalase. The formation of carnitine was high with ascorbate and isoascorbate, lower with reduced 2,6-dichlorophenolindophenol and 2-amino-5,6-dimethyl-4-hydroxy-5,6,7, Metrahydropteridine, and still lower with the other reductants which were tested.
Catalase-The stimulatory effect of catalase on the rate of carnitine formation was the same with 2-ketoglutarate as cofactor as with the NADPH-isocitrate dehydrogenase system (6). An addition of at least 0.3 mg of catalase per ml was required for maximal rate of carnitine formation.
A low rate of carnitine formation was observed when the enzyme had been Effect of sulfhydryl reagents on formation of carnitine from r-butyrobetaine The purified protein fraction (0.5 mg) was incubated for 20 min at 37" with the inhibitor in 0.3 ml of 0.15 M potassium phosphate buffer, pH 6.5. [Methyl-W,Jy-butyrobetaine, 2-ketoglutarate, ferrous ion, ascorbate, and catalase were then added, and the incubation (0.8 ml) was continued for 1 hour at 37". For details In the previous studies of the hydroxylation of y-butyrobetaine to carnitine in rat liver (6) some results were obtained for which no explanation could be offered, such as a stimulatory effect of microsomes, a dual requirement for a reductant, wiz. ascorbate and a NADPH-regenerating system, as well as a stimulating effect of catalase.
We had also observed that NADP+ was as effective as NADPH when crude extracts of acetone-dried rat liver were used as sources of enzyme and ascorbate, ferrous ion, and fumarate were the other cofactors (7). The results now reported indicate a specific requirement for 2-ketoglutarate in y-butyrobetaine hydroxylation in the rat liver, and that the observed stimulation by NADPH, isocitrate, and isocitrate dehydrogenase is related to the formation of 2-ketoglutarate.
As microsomes had negligible effect in the incubations with 2-ketoglutarate, they probably stimulate the formation of carnitine in incubations with the isocitrate dehydrogenase system by oxidizing NADPH thereby causing an increased formation of 2-ketoglutarate.
The fact that none of a series of organic acids could replace 2-ketoglutarate and the disappearance of hydroxylating activity in a 100,000 X g supernatant fraction of a rat liver homogenate after preliminary treatment with 2-ketoglutarate aminating systems may be taken as evidence that 2-ketoglutarate is the cofactor which participates in uivo in the hydroxylation reaction.  7. Relation between the enzymic activity (r-butyrobetaine hydroxylation) and the pH value in the incubations. See "Assay" for details.
inhibitors of the enzymic activity of the partially purified protein fraction.
Organic mercurials and alkylating agents were effective inhibitors, whereas arsenicals were less effective under the incubation conditions used (Table V). No study has been made of the time course of the inhibition. However, the inhibition was less pronounced when the inhibitors were added to the enzyme simultaneously with the cofactors than when they had been previously incubated with the enzyme for 20 min at 37". E$ect of pH Value-The rate of carnitine formation was approximately the same within a fairly wide pH interval, viz. 6.4 to 8.0 in the previous study of the y-butyrobetaine hydroxylase from rat liver, when the NADPH-isocitrate dehydrogenase system was used as one of the cofactors (14). The pH dependence was therefore studied with 2-ketoglutarate as cofactor instead of the NADPH-isocitrate dehydrogenase system. Fig. 7 shows that under these conditions the rate of carnitine formation was highest when the pH value was about 6.7.

2-Ketoglutarate
might be bound to the enzyme by way of the carbonyl group and the y-carboxylate group, as succinic semialdehyde was inhibitory and 2-ketovalerate was not. The other aldehydes which were tested had little or no inhibitory activity. However, kinetic studies with the inhibitors are required to solve this problem.
The stoichiometric relationship between degradation of 2ketoglutarate and hydroxylation of y-butyrobetaine suggest an intimate coupling between the oxidative decarboxylation of 2ketoglutarate and the formation of the hydroxylated product. Similar results were obtained with y-butyrobetaine hydroxylase from a Pseudomonas strain (21), which has the same cofactor requirements as the enzyme from rat liver (22, 23). Probably 2-ketoglutarate acts as the specific reductant of oxygen. The electrons may be transferred either directly to oxygen concomitantly with the decarboxylation or via an intermediate electron carrier.
In the latter case the reaction would be similar to that catalyzed by the 2-ketoglutarate dehydrogenase and pyruvate dehydrogenase complexes in the mitochondria in which thiamine pyrophosphate and lipoic acid are cofactors. However, such a mechanism for y-butyrobetaine hydroxylase appears less probable as no enzymic decarboxylation of 2ketoglutarate could be demonstrated in the absence of ybutyrobetaine, whereas in the thiamine pyrophosphate-dependent decarboxylations, degradation of the 2-keto acid may be noted also in the absence of oxidants (24,25).
Recently (17), we proposed a reaction mechanism for 2-ketoglutarate-requiring hydroxylases (Scheme 1) according to which the anion of the substrate to be hydroxylated (I in Scheme 1) is attacked by a positively charged ferrous ion-oxygen complex, after which dissociation of ferrous ion occurs simultaneously with a nucleophilic attack of the hydroperoxide anion (II) on the 2-carbon atom in free or enzyme-bound 2-ketoglutarate. Rearrangement of the resulting peroxide (III) would result in decarboxylation and formation of succinate and the anion of the hydroxylated