The Enzyme-Enzyme Complex of Transaminase and Glutamate Dehydrogenase *

Glutamate dehydrogenase decreases the distribution COefficient of glutamate oxalacetate transaminase in Sephadex G-ZOO. This is consistent with previous results which suggested that a complex is formed between these two enzymes. These gel filtration as well as kinetic experiments suggest that transaminase can react with monomeric but not polymer forms of glutamate dehydrogenase. When the levels of both mitochondrial enzymes are too low to form a complex, there is little TPNH and NH4+ generated by the combined aspartate aminotransferase and glutamate dehydrogenase reactions. This is because oxalacetate is a potent product inhibitor of transaminase, the reaction with glutamate and glutamate dehydrogenase is slow and ol-ketoglutarate, oxalacetate, and aspartate all inhibit this latter reactions. When the levels of both enzymes are sufficiently high to form a complex, asparate can be dehydrogenated quite rapidly even in the absence of oc-ketoglutarate. Furthermore, the aspartate dehydrogenase reaction catalyzed by the enzyme-enzyme complex is not markedly inhibited by oxalacetate, or activated by oc-ketoglutarate, and can take place even in the presence of rather high levels of glutamate. Kinetic and gel filtration experiments suggest that the dissociation constant of the enzyme-enzyme complex is considerably lower than that of these substrates for the free enzymes. These results suggest that an important physiological function of the enzyme-enzyme complex is to catalyze the asparatate dehydrogenase reaction in organs as brain, liver, and kidney, where the mitochondrial levels of these enzymes are sufficiently high to form a complex. An advantage of catalysis by the complex over transamination with aspartate followed by dehydrogenation of glutamate is that the complex is not markedly inhibited by low levels of oxalacetate. When tyrosine is the substrate and the levels of these two enzymes are too low to form a significant amount of complex, the tyrosine aminotransferase and glutamate dehydrogenase

reactions are coupled.
This can occur with tyrosine but not aspartate because hydroxyphenylpyruvate is not a potent inhibitor.
However, when the levels of both enzymes are sufficiently high to form a significant amount of the enzymeenzyme complex, the tyrosine dehydrogenase reaction can be catalyzed by this complex and again oc-ketoglutarate has no effect on the reaction. Similar relationships occur when phenylalanine is the substrate. Thus, the enzyme-enzyme complex can facilitate dehydrogenation of these amino acids which do not react with glutamate dehydrogenase.
Several experiments tlctnonstratc that this reaction results from formation of an ctlzyme-cnzytnc comples and is tlot tnediated by kcto or amino acids bountl to cithcr cnzymc (l-4) and that, in this reaction, glutatnatc tlrltgtlrogetiase reacts with transaminase bound pyritlosal phosphate (l-4).
That is, in this reactSion the transamitiasc is not, functioning as a catalyst but &livers the actual substrate, pgridosal phosphate, to the active site of glutamate tlchydrogetiase.
This reaction can take place in the abscncc of kcto or amino acid. However, if osalacetate, a poor substrate of glutamate tlchytlrogenas~~, is ad&l, this keto acid is reductively aminated by the combination of the cnzytnc-enzyme reaction plus the transaminasc onc-half reaction.
TPNH + NH,+ + pyridoxal-I'-trnnsaminase ti TPN+ + pS'ridoxRmitte-P-trarlsaminase + II,0 Pyridoxamine-P-transaminase + oxalacctate ti aspartate + pyridoxal-P-transaminase This manuscript presents adtlit~ional studies and evidence for the formation of the enzymecnzytne complex. Also inclutletl arc kinetic studies of amino acid tlrhytlrogenase reactions catalyzed by the bovine liver tnitochondrial glutamate dehytlrogcnasc-glutarnatc osalacetntc tratisaniitiase complex. Thus these experiments arc pcrformetl with two mitochondrial cnzymes both isolatctl from the same organ. METHODS AND MATERIALS were incubated in 0.025 M sodium arsenate-O. Concentration-The con-15 centration of transaminase and glutamate dehydrogenase is expressed in units of micromolar peptide chain. In the case of transaminase this was measured and calculated with previously described methods (1,2,10,11). The concentration of glutamate IO dehydrogenase peptide chains was calculated from spectrophoto-/v metric measurements using 0.97 mg ml-l cm-l as the extinction coefficient (12) and 5.6 X lo4 as the molecular weight of a peptide chain (13,14).
Measurements of Products-The concentration of oxidized or 5 reduced pyridine nucleotides and pyridoxamine or pyridoxal phosphate produced in reactions was measured as previously described (l-3).
The rate of oxalacetate production was measured spectrophotometrically at 260 to 280 nm (15) or in a coupled assay with DPNH plus malate dehydrogenase (16 tivity or the addition of mitochondrial glutamate oxalacetate transaminase results in tyrosine or phenylalanine dehydrogenase activity (Fig. I). These transaminases do not have amino acid dehydrogenase activity.
The effect of these transaminases is not due to protection of glutamate dehydrogenase against, inactivation since this enzyme is stable in these experiments.
These amino acid dehydrogenase reactions take place nith low levels of transaminase (Fig. I), do not require added keto acids, and, therefore, apparently are mediated by the previously proposed enzyme-enzyme complex between transaminase and glutamate dehydrogenase.
An alternative explanation is that these reactions are mediated by oc-kctoglutarate or glutamate impurities in the system. These arc the only known substrates sufficiently reactive with all four of these enzymes to conceivably play this role. Furthermore, low (micromolar) levels of glutamate-oxalacetate transaminase can increase the rate of reductive amination of oxalacetate or dehydrogenation of aspartate when the level of these substrates are high (1 to 10 mM). Similarly, low levels of glutamate pyruvate transaminase can increase t,he rate of reductive amination of pyruvate or dehydrogenation of alanine when the levels of these substrates are high (1 to 10 mM). To explain such increases on the basis of substrate contamination would require that micromolar levels of these transaminases would contain millimolar amounts of these substrates, and this is not the case ( Fig. 1 and Refs. 1 and 2).
Estimates of Levels of Substrate--If the two enzymes are assumed to be coupled through the presence of contaminant ol-ketoglutarate or glutamate in the experiments of Fig. 1, how much of these would be required to explain the observed rate? To estimate this, assays were performed with a range of these substiates (10 to 100 PM) and wit,h aspartate (10 m&l) plus cytoplasmic glutamate oxalacetate transaminase alone or with TPN (1.0 mM) plus glutamate dchydrogenase alone. The initial velocity in the presence of either of these enzymes alone with very low levels of these substrates then was calculated by extrapolation from the standard curve t,hat was obtained. These calculations are apparently valid since the observed velocities were proportional to enzyme concentration; that is, the concentration of free and added substrates are identical (because the dissociation constants of bot,h substrates for both enzymes are quite high).
Furthermore, the measured velocities were observed to decrease linearly as the concentration of substrate remaining in the reaction decreased (because the concentration of substrate was much lower than the K,,).
The rate of TPNH production observed in the presence of TPN, aspartate, cytoplasmic glutamate oxalacetate transaminase (0.2 PM), and glutamate dehydrogenase (7.2 PM), conditions similar to those used in Fig. IB, is (after correcting for the slow rate of aspartat,e oxidation catalyzed by glutamate dehydrogenase in the absence of transaminase) equal to that expected in the transaminase reaction with aspartate (as assayed with the coupled malate dehydrogenase system) and 0.13 I.~M a-ketoglutarate or the glutamate dehydrogenase reaction with TPN and 4 PM glutamate.
Therefore, if the enzymeenzyme reaction is actually due to recycling of these substrates between the two enzymes, then the level of either of these substrates would have to be at least 4 PM or the molar ratio of these substrates to transaminase or glutamate dchydrogenase peptide chains would have to be respectively 20 or 0.6. This is not consistent with experiments performed with the reverse reaction which revealed that the molar ratio of these substrates to transaminase or glutamate dehydrogenase could not be higher than 0.3 or 0.08. This is, respectively, 66-and 'i-fold lower than that required to explain the rate observed with TPN, aspartate, and both enzymes.
Similarly, it was found that, in the presence of the mitochondrial transaminase (0.25 phf), TPN (1 .O mM), aspartate (10 mM), and glutamate dehydrogenase (10 PM), the rate of TPNH production (after subtracting the slow rate in the absence of transaminase (see Fig. 1)) was equivalent to that expected in this transaminase react,ion (as assayed with the coupled malatc dehydrogenase system) with aspartate and 1.0 pM cr-ketoglutarate or the glutamate dehydrogenase reaction with TPN and 4 /.LM glutamate.
Thus, if this reaction is mediated by oc-ketoglutarate or glutamate associated with this transaminase, the molar ratio of these substrates to transaminase peptide chains would have to be 16. Again, this is not consistent with experiments performed with the reverse reaction which revealed that the molar ratio of these substrates to transaminase could not be higher than about 0.03 or about 500-fold lower than the value required to explain the reaction with TPN and aspartate.
Thus, the reductive amination of oxalacetate and dehydrogenation of aspartat,e cannot both be explained on the basis of a constant ratio of these substrates to these enzymes. Also, a high level of either of these substrates would be required to explain the rate observed with TPN, aspartate, and these enzymes.
Previous experiments have ruled out the possibility that the ratio of cr-ketoglutarate or glutamate to the cytoplasmic transaminase is 20 (3). Similar experiments were performed which ruled out the possibility that this ratio is 16 in the case of the mitochondrial t.ransaminase.
In addition neither cr-ketoglutarate not glutamate are detected when 30 to 60 nmoles of either of the transaminases or glutamate dehydrogenase are deproteinized and the centrifuged protein-free supernatant solution is assayed with the amino acid analyzer, with DPNH, and NHhf plus native glutamate dehydrogenase, or with 2,4-dinitrophenylhydrazine.
If these substrates were associated with these enzymes in levels even considerably lower than required to explain the rate observed with TPN and aspartatc, then they would have been detected in these assays. Also when tyrosine was incubated with the pyridoxal phosphate form of the mitochondrial enzyme it was found that pyridoxamine phosphate and hydroxyphenylpyruvate were produced in equal amounts, and glutamate was not detected when 20 nmoles of this tyrosine treated transaminase was deproteinized and applied to the amino acid analyzer.
Therefore, there is no evidence that tyrosine is transaminating with endogenous a-ketoglutarate. The possibility that these substrates are associated with glutamate dehydrogenase was also eliminated by finding that high levels of this enzyme do not oxidize TPNH (in the presence of NH4f) or reduce TPN.
Also incubating this enzyme with 50 mM phenylhydrazine for 1 hour followed by dialysis does not alter its ability to react with transaminase.
The above and other experiments (l-4) rather conclusively rule out the possibility that these reactions are mediated by contaminating amounts of a-kctoglutarate or glutamate associated with the enzymes. It is also unlikely that these substrates are contaminants of the reagents. First they would have to be rather ubiquitous contaminants since enzyme-enzyme interaction occurs with TPNH, DPNH, DI'N, and TPN with or without several different keto or amino acids (l-4) and in all buffers tested (1). Also it would be difficult to explain why the reaction with TPN plus aspartatc would be saturated with such low levels of transaminase (0.2 pM, see Fig. 1) on the basis of impurities in the reagents. Furthermore, assays performed with high levels of these reagents similar t,o those described for enzymes, i.e. amino acid analysis and reactivity with glutamate dehydrogenase or 2,4-dinitrophenylhydrazine, were all negative.
Enzyme-Enzyme Complex-Glutamate dehydrogenase, but not other proteins tested, decreases the distribution coefficient of transaminase in Sephadex G-200 (Table I). This is not because t'his enzyme protects transaminase against inactivation or activates t.ransaminase.
Control experiments performed without Sephadex revealed that transaminase is completely stable under t.hese condit.ions and when the enzymes are present in the low levels used in assays glutamate dehydrogenase does not alter transaminase activity. Furthermore, none of the substrates or coenzymes used increased the distribution coefficient of transaminase in the absence of glutamate dehydrogenase.
If these compounds were inactivating transaminase and glutamate dehydrogenase was protecting, then one would espect the apparent distribution coefficient to be increased in the presence of these compounds and absence of glutamate dehydrogenase.
Also, the amount of transaminase in the aqueous phase was constant for 2 to 8 hours. Thus it is unlikely that transaminase is labile for 2 hours and stable for the remaining 6 hours. In addition at the end of these gel filtration experiments gels containing transaminase were transferred (by washing and stirring with buffer) into small glass columns.
The gels then mere eluted until transaminase Jyas no longer recovered in the eluate. About SO70 of the transaminase units were recovered.
This would be the expected recovery (within experimental error) if there were no loss of transaminase activity during the course of these experiments.
The distribution coefficient of transaminase in the presence of DPK plus aspartate or DPNH plus NH4+ and glutamate   dehydrogenase is so low that the only possible conclusion is that under these conditions glutamate dehydrogenase decreases the ability of transaminase to penetrate Sephadex G-200. Such a complex was proposed from the results of past kinetic experiments (l-4). This complex, which would have a molecular weight of at least 4 x 105, would be excluded from Sephadex G-200 unlike free transaminase, which has a molecular weight of 9 x 104 (13,14,21).
The fraction of transaminase (X) bound to glutamate dehydrogenase in these experiments cm be estimated with the use of Equation 1: where E is the concentration of transaminase in the aqueous phase, ET is the total amount of transaminase added, V', is the volume of the aqueous phase, Vt is the total volume, and Ku is the distribution coefficient of transaminase in the absence of glutamate dehydrogenase (see "Methods and Materials").
Since the concentration of glutamate dehydrogenase used was not the same in all experiments, the results are also espressed (Table  I) as an apparent equilibrium constant, K, which is the concentration of glutamate dehydrogenase in the aqueous phase (glutamate dehydrogenase does not penetrate Sephadex G-200) times the ratio of the fraction of transaminase free to the fraction of transaminase bound.
Binding of transaminase to glutamate dehydrogenase is maximal when either DPNH plus NH4+ or DPN plus aspartate are present (Table I). There is much less binding in the presence of DPN plus NH .++ than with DPN plus aspartate.
In erperimerits with DPNH, NH4+, and glutamate dehydrogenase binding of transaminase reaches a maximum when the level of glutamate dehydrogenase is 10 pM (Fig. 2). That is, there is no additional binding when higher levels of glutamate dehydrogenase are added.
Reaction with Aspartate-A concept of the physiological role of these enzymes in mitochondria is that aspartate reacts with Lu-ketoglutarate to produce osalacetate plus glutamate. Glu The rate of oxalacetate formation in the presence of 10 mM aspartate and 100 FM cu-ketoglutarate is shown in Curves A and U. The results shown in Curve A were measured with the coupled DPNH plus malate dehydrogenase assay while those shown in Curve B were measured directly in the absence of DPNH plus malate dehydrogenase.
The rate of TPNH production in the presence of 10 mM aspartate, 8 jaM glutamate dehydrogenase, 1.0 mM TPN, and presence or absence of 100 j&M a-ketoglutarate is shown, respectively, in Curves C and D. The other experimental conditions are described in the legend to Fig. 1. FIG. 4 (cenler). Plot of velocity (in nanomoles per min per ml) versus keto acid concentration.
The effect of or-ketoglutarate on the rate of TPNH production in the presence of 1.0 mM TPN, 10 mM aspartate, 8 PM glutamate dehydrogenase, and 0.2 or 0.012 tamate is then dehydrogenated via the glutamate dehydrogenase reaction so the over-all net effect is an aspartate dehydrogenase reaction.
However, in the presence of TPN, aspartate, a-ketoglutarate, high levels of glutamate dehydrogenase, and levels of transaminase which are low but sufficiently high to produce glutamate at a rapid initial rate, t.here is little coupling between the two enzymes and TPNH is produced quite slowly (Fig. 3, Curve C). This in part is because the oxalacetate generated in the first step is a potent product inhibitor of transaminase (Fig.  3, Curves A and B; Refs. 22 and 23). Thus when the level of transaminase is low, the slow dehydrogenation of aspartate by glutamate dehydrogenase is the main TPNH producing reaction and this reaction is inhibited by a-ketoglutarate (Fig. 3, Curve C, and Fig. 4, Curve c).
If the concentration of both enzymes are high, then TPNH and oxalacetate can be rapidly produced by the enzyme-enzyme complex even in the absence of a-ketoglutarate (Fig. 3, Curve D). Furthermore, under these conditions, oc-ketoglutarate has little effect on the rate of TPNH production especially if the level of glutamate dehydrogenase is quite high. That is, increasing the level of glutamate dehydrogenase in the presence of a high level of transaminase decreases the effect of cu-ketoglutarate (Fig. 5). Also, when the levels of both enzymes are high so that the reaction can be catalyzed by the complex, oxalacetate is not a potent inhibitor as it is in the reaction with a-ketoglutarate and transaminase (Fig. 4)  the equilibrium state of the glutamate dehydrogenase reaction is essentially equal to the amount of TPNH produced in the absence of glutamate (Fig. 6).
Reaction with Tyrosine-When the concentrations of glutamate dehydrogenase and transaminase are low, DPN plus glutamate dehydrogenase can serve as a coupled assay of the tyrosine-cu-ketoglutarate amino transferase reaction (Fig. 1,  Curve A). Thus the rate of this reaction is: (a) the same in either the presence or absence of DPN plus glutamate dehydrogenase; (b) directly proportional to transaminase concentration (Fig. 1, Curve A) ; and (c) independent of glutamate dehydrogenase concentration over a range from 1 to 14 pM (Fig.  7, Curve A). Furthermore, the apparent K, of a-ketoglutarate is in the same range whether this reaction is measured directly or in the coupled assay with DPN plus glutamate dehydrogenase.' Apparently, glutamate dehydrogenase can assay the tyrosine but not the aspartate amino transferase reaction because hydroxyphenylpyruvate unlike oxalacetate is not a potent inhibitor of transaminase.
When the level of glutamate dehydrogenase is high so that DPNH and hydroxyphenylpyruvate can be produced via the enzyme-enzyme complex, then ar-ketoglutarate has no effect 1 The K,, of a-ketoglutarate or oxalacetate cannot be measured accurately in the absence of DPN plus glutamate dehydrogenase. However, it can be stated that these values are quite low since in the tyrosine transferase reaction velocity is independent of a-ketoglutarate or oxalacetate concentration over a range of these substrates from 0.1 to 1.0 mM.  (Fig. 1, Curve B), (5) is proportional to glutamate dehydrogenase concentration over the range from 1 to 10 FM (Fig. 7, Curve B), (c) is inhibited by ADP (Fig. 8, Curve B), and (d) is activated by GTP (Table II).
When the level of glutamate dehydrogenase is low and this enzyme is reacting with glutamate rather than with the pyridoxamine phosphate form of the transaminase, GTP is an inhibitor (Table II).
Reaction zuifh Phen?llalanine-Reactiolls with phenylalanine are similar but slower than those with tyrosine ( Figs. 1 and 7). Again, a-ketoglutarate has no effect on the tlehydrogenase reaction catalyzed by the enzyme-enzyme complex Reaction with Pyridoxal Phosphate and Pyridozamine Phosphate-These coenzymes are such poor substrates of glutamate dehydrogcnase that a kinetic study is hardly feasible. For example, if pyridosamine phosphate (0.1 mM) is incubated with DPN or TPN (1.0 HIM) and glutamate dehydrogenase (0.85 mg per ml), the initial velocity is hardly detectable.
However, the fact that both pyridoxal phosphate and pyridoxaminc phosphate are substrates for glutamate dehydrogenase can be confirmed by carrying out the incubation with higher levels of these coenzymes. Fractional chromatography of the incubation on DEAE-Sephadex gave fractions which eluted as and had an absorption spectrum like the suspected product.
The identity of the product was confirmed in each case by finding that it could reconstitute the transaminase apoenzyme.  III addition the fact that WI' activates and XI>l' inhibits the tyrosine dchytllogcnasr wartion cntalyzcd by the cnzymc-cnzymc system also csscntially lulrs out the possibility that this reaction is mcdiatctl by wcyclitlg of a-kttoglutaratc and glutamate between the two c~lzymos. If this wre the cast tllril GTI' would inhibit aiitl AD1 xvoriltl activate (24). \\Mc these transaminascs appawt~tly possess a kcto group which can osidizc DI'NII (but not 'I'I'SI I) (3), this group tlocs not react with glutamate dehytl~ogr~iasc siiicc rcducitig it with SaHII., or rcactiiig it, with l)llcil~lli~tl~~~xi~i(~ tlocs not inhibit the abilitjy of transnminase to react with glutamate tlcl~ydroge~~asc (4). Srvrral csperimciits suggrst that, I~~I~I~ the levrl of glutwmatc drhytlrogcnasc is high (8 par This is consistent nith the fact that GTP which dissociates glutamate dchydrogenasr (24) activates reactions with tyrosinc and the enzyme-eiizymc romplcs while AI)l' which facilitates association inhibits both the tyrosiiie and a:);,artate deliydrogci~ase reactions rata,lyzed by the complcs. The activating effect of GTI' is not large. l~Iowcvcr, if GTI' (0.1 m3I) is added to the assays pcrformod with aspartatc tlcscribed in the lcge~~l to Fig. 1, thr weight avcragc molecular wiglit. is decreased only from about 1.5 to 1.1 X lo6 (as estimatcd by light scattering).
Thus, undrr thcsc conditions GTI' does not markedly dissociate glutamate dehydrogenase. Furthermore, it is conceivable that GTI is not a potcnt activator of the tyrosine dchydrogrnasc traction bccausc, xyhilc it increases the concentrations of monomrric forms of glutamate drl~~drogeiiase, it might also inhibit the reaction ratalyzctl by thcsc forms. Thus, the slight actiratioll could bc the net result of two opposing effects.
Other esperimcnts also suggest that, only monomeric forms of glutamate dcliydrogcnasc wart with trausaminasc. For example, in the presence of rrducctl pyridine nuclcotidrs, plus NIId+, a level of glutamate tleliydlogcnasc is reached xhcre adding more enzyme neither illcrrascs the rate of conversion of transaminasc from the pyridosal to pyridoraminc phosphate form 1x01' the amount of traiisamiiiasc boulld. This suggests that when glutamate dehgdroge~lase rrachcs a ccrtaiu level it is mainly po1ymerized and addiiig mow enzyme does not proportionally increase the conccntratioll of monomers. Furthermore, a con~plcs between these two enzymes has not been detected in sedimentation esperimellts where the levels of both enzymes must bc high for visualization.
The csplanatioii for this may bc that in these experiments glutamate dchydrogenase is mainly polymeric and consequently oiily a small fraction of this enzyme can form a complcs with transaminase.
Also, siucc transaminase only forms a complcs with the small fraction of glutamate dchydrogenasc molcculcs which arc monomeric, most of the transaminasc molecules are not bound. COIISC-qucntly, the two riizyincs arc pwscnt primarily as free rnzymes in the prcsriice of a small, iiontlrtrctablr amount of ciizyme-c~lzgmc complcs.
Thcrcforc, in thcsc csperimcnts the amount of rnzyme-enzyme coml~lcs is quite significant compared with the total amount of transaminasc atldcd and thus the complcs can be detected.
This is iii spite of the fart that pyridosal phospliatc can be reducctl iiolrcilz~iiinticall~ by various 1 ,4-dihydropyritliiics (26). It is possiblr, as suggcstrtl by u~~l~ublishctl cspcrimciits, that the transamiiiase molcculc is nrccssay for significant binding of pyridosal or pyridosamine phosphate to the active site of glutamate tlchydrogrxw.
The cnzgme-cnzymc complcs can bc formed iii the prcscncc of high levels of cu-kctoglutaratr, osalacctatc, or glutamate. 011 the basis of gel filtration osprrimcnts, the dissoriation constant of the riizgmc-cnzymr complex is apparently lower than that) of these substrates for cithcr of the free enzgmcs (16, 22). A possible physiological role of the e~zymc-cnzymc complcs would bc to catalyze amino arid dehydrogrnase reactions with amino acids which arc not very rcactivc with glutamate dehydrogcnasc but arc rcactivc with a transaminasc.
The advantage of cntalyzing the aspartate tlel~ytlrogenasc waction by the enzymc-enzyme complcs over wrycling glutamate and Lu-ketoglutaratc bctxccn the two enzymes is that the former is uot markctlly inhibitctl by low lcvcls of osalacctatc.
The enzyme-euzymc complcs could bc important in tyrosinc metabolism because: (n) the rate of conversion of tyrosine to llS.tl~os~l~lleii~ll~~r~i\~atc is in the same range rcgartllcss of n-hcthcr I)I'N plus glutamate tlchydrogcnasc or saturating levels of osalacetntc or oc-kctoglutarate rccyclc the pyridosamine phosphate form of the transaminasc; (b) both cnzymrs are prcsent in high levels in brain and livrr mitochondria (27)(28)(29); (c) glutamate osalacctatc transamiilasc is apparently the only bovine liver mitocholldrial protcill that can transaminatc tyrosinc (since the ratio of nspartatc to tyrosinc amiiiotraiisfr~nsc activity is coiistant through purification).
Thus it seems possible that, in brain, for csample, thcsc reactions could bc rapid compared with known slow strps ill tyrosine metabolism as tyrosinc hgdrosylase (30). This could bc rrlatcd to the fact that some tranquilizers are potctlt i7l viva n11t1 in vilro inhibitors of glutamate dchydrogcllasc (29, 31, 32) but do not inhibit reactions catalyzed by the enzyme-enzgmc complcs (33).
The levels of thcsc two ciizgmes are suficicntly high in viva to interact.
Both enzymes arc rather uniformly distributed in mitocholldria (34,35). It has been estimated that the lcvcl of glutamate dehydrogcnasc ill bovine liver mitochondria is 1 to 1.25 mg per ml or 20 par (36). On the basis of the ratio of specific activity in ow mitochontlrial estrarts to that of the pure enzymes, the lcvcl of glutamate osalacetate transamiiiase is about 3-fold lower than that of glutamata tlehydrogrnasc (5, 16). Thus, the lcvcl of transaminase would be about 6 pin. These levels are even higher than required for interaction in in z&o experiments.