Benzoyl-coenzyme A:glycine N-acyltransferase and phenylacetyl-coenzyme A:glycine N-acyltransferase from bovine liver mitochondria. Purification and characterization.

Two closely related acyl-CoA:amino acid N-acyl-transferases were purified to near-homogeneity from preparations of bovine liver mitochondria. Each enzyme consisted of a single polypeptide chain with a molecular weight near 33,000. One transferase was specific for benzoyl-CoA, salicyl-CoA, and certain short straight and branched chain fatty acyl-CoA esters as substrates while the other enzyme specifically used either phenylacetyl-CoA or indoleacetyl-CoA. Acyl-CoA substrates for one transferase inhibited the other. Glycine was the preferred acyl acceptor for both enzymes but either L-asparagine or L-glutamine also served. Peptide products for each transferase were identified by mass spectrometry. Enzymatic cleavage of acyl-CoA was stoichiometric with release of thiol and formation of peptide product. Apparent Km values were low for the preferred acyl-CoA substrates relative to the amino acid acceptors (10(-5) M range compared to greater than 10(-3) M). Both enzymes were inhibited by high nonphysiological concentrations of certain divalent cations (Mg2+, Ni2+, and Zn2+). In contrast to benzoyltransferase, phenylacetyltransferase was sensitive to inhibition by either 10(-4) M 5,5'-dithiobis(2-nitrobenzoate) or 10(-5) M p-chloromercuribenzoate; 10(-4) M phenylacetyl-CoA partially protected phenylacetyltransferase against 5,5'-dithiobis(2-nitrobenzoate) inactivation but 10(-1) M glycine did not. For activity, phenylacetyltransferase required addition of certain monovalent cations (K+, Rb+, Na+, Li+, Cs+, or (NH4)+) to the assay system but benzoyltransferase did not. Preliminary kinetic studies of both transferases were consistent with a sequential reaction mechanism in which the acyl-CoA substrate adds to the enzyme first, glycine adds before CoA leaves, and the peptide product dissociates last.

Mammals conjugate a variety of aliphatic and aromatic monocarboxylic acids with amino acids and the resulting peptides appear as excretory products in the urine and bile. This process is physiologically important in detoxification and shows species specificity with respect to both the acyl and amino acid components of the peptides (1,2). Although such specificity is probably conferred by a class of enzymes termed acyl-CoA:amino acid N-acyltransferases only one of these proteins has been purified to near-homogeneity (3) and much needs to be learned about their number, properties, intracellular loci, genetics, and evolution.
Schachter and Taggart first discovered acyl-CoA:amino acid N-acyltransferase activity in preparations of pig kidney cortex (4). Later, they showed that a partially purified enzyme from bovine liver mitochondria catalyzed acyl transfer from a wide variety of aliphatic and aromatic acyl-CoA substrates specifically to glycine (5). The activity of the bovine preparation with both benzoyl-CoA and phenylacetyl-CoA was hard to reconcile with the recent findings of Webster and co-workers who separated and partially purified the benzoyl-and phenylacetyltransferase activities in mitochondrial preparations from rhesus monkey and human liver (6). One enzyme fraction catalyzed benzoyl transfer from benzoyl-CoA specifically to glycine while the other catalyzed phenylacetyl transfer from phenylacetyl-CoA specifically to L-glutamine. The acyl-CoA substrates for one transferase fraction inhibited activity of the other.
We now find that bovine liver mitochondria actually contain two closely related N-acyltransferases with acyl-CoA substrate/inhibitor specificities resembling those described above for rhesus monkey and man (6). Both a benzoyltransferase and a phenylacetyltransferase have been separated, purified to near-homogeneity, and characterized. Similarities and differences between these two enzymes are the subject of this report. Enzyme Purification and Stability-Benzoyltransferase was purified about 24-fold from a supernatant solution derived from frozen and thawed bovine liver mitochondria (Table IS); the corresponding purification for phenylacetyltransferase was approximately 940-fold (Table IIS). Although the crude mitochondrial preparation initially contained nearly 19 times as much benzoyltransferase as phenylacetyltransferase activity, the specific activities of the respectively purified enzymes were 24 and 51 pmol of acyl-CoA substrate cleaved/min at 30"C/mg of protein. The purification procedures for both transferases were quite similar except for the final separation by hydroxylapatite chromatography (Fig. 1, Tables IS and  IIS).
Both enzymes retained about 80% of their activities after 1 to 3 weeks storage at 4°C and the benzoyltransferase retained more than half of its activity after several months storage at -70°C. Phenylacetyltransferase fractions from the hydroxylapatite column were inactivated by dialysis or concentration by ultrafiltration.
Physical Properties-By SDS2-disc gel electrophoresis, each enzyme appeared as a single band located between carbonic anhydrase, 29,000, and glyceraldehyde-3-phosphate dehydrogenase, 36,000 (Fig. 2). The phenylacetyltransferase did not separate completely from the higher molecular weight protein marker whereas the benzoyltransferase did. Together the two enzymes produced a single broad band. By conventional disc gel electrophoresis, phenylacetyltransferase appeared as a heavy band closely preceded by a very light band; the heavy band was located at the same position where enzymatic activity was found (Fig. 1s). Benzoyltransferase showed three protein bands which corresponded to three peaks of enzymatic activity (Fig. 2s); the anomalous electrophoretic behavior of this enzyme was not investigated further.
Both enzymes had molecular weights near 33,000 (Table  111s). In contrast to SDS-disc gel electrophoresis, gel filtration gave a slightly higher molecular weight for benzoyltransferase than for phenylacetyltransferase.
Both enzymes had identical molecular weights as judged by sucrose density gradient centrifugation.
Agreement of the SDS-gel electrophoresis results with those obtained by the other two methods indicates that each enzyme consists of only a single polypeptide chain. Reaction Requirements, Peptide Products, and Stoichiometry-Formation of benzoylglycine (hippurate), benzoylasparagine, and benzoylglutamine from benzoyl-CoA depended on active enzyme and the appropriate amino acid acceptor; comparable results were obtained with phenylacetyltransferase (Table IVS). Both enzymes were stereospecific for the L forms of asparagine and glutamine. When enzymatically formed peptides were isolated by high pressure liquid chromatography, mass spectra of their methyl ester derivatives resembled those recorded for the corresponding chemically synthesized standards." Balance studies of both transferase reactions (  Table  IIS, Footnote 6, for column conditions. Fractions were assayed for benzoyltransferase activity, phenylacetyltransferase activity, and protein as described under "Materials and Methods" in the miniprint. tonyl, and tiglyl transfer from the corresponding acyl-CoA to glycine (Table VIS). Malonyl-CoA did not serve as a substrate and only marginal activity was observed with methylmalonyl-CoA. The hydroxylapatite fraction having the highest transferase activity with benzoyl-CoA also showed the highest activity with the other active acyl-CoA substrates; this suggested that all activities were derived from a single enzyme. Phenylacetyltransferase utilized phenylacetyl-CoA and indoleacetyl-CoA as arylacetyl-CoA substrates; relative nearmaximal rates were about 2.5 to 1 at 0.1 UIM acyl-CoA and 0.2 M glycine. Acyl-CoA substrates for one transferase did not serve as substrates for the other but instead those tested acted as competitive inhibitors with respect to the preferred acyl-CoA substrate (Fig. 3s). At high concentrations of glycine, apparent K, values for the preferred acyl-CoA substrates of both enzymes were in the 1O-5 range while they were even higher for nonpreferred acyl-CoA substrates of benzoyltransferase (Table VIIS).
Glycine was the preferred acyl acceptor for both enzymes but L-asparagine and L-glutamine were also active. With different preparations of benzoyltransferase, L-asparagine at 0.1 M gave 10 to 20% of the rate observed with 0.1 M glycine and 0.1 UIM benzoyl-CoA; the corresponding rate with 0.1 M Lglutamine was 5 to 7%. The apparent K, values for L-asparagine and L-glutamine were appreciably higher than that of -3 mM found for glycine; substitution of these acceptors for glycine also raised the apparent K, for benzoyl-CoA (Table  VIIS). Phenylacetyltransferase also had higher apparent K, values for L-asparagine and L-glutamine than for glycine even though its apparent K,,, for glycine (20 mu) was higher than that of benzoyltransferase (Table VIIS). Other L-amino acids listed in the supplement under "Substrate Specificity," as well as n-alanine, D-ghtamine, methylamine, N,N-dimethylglytine, and aminomethylphosphonate, did not serve as acyl acceptors; the last three compounds acted as weak inhibitors of benzoyltransferase with respect to glycine (Fig. 4s).  Benzoyltransferase displayed maximal activity in the presence of the assay buffer (25 mu Tris-Cl, pH 8) whereas phenylacetyltransferase was nearly inactive (Table I, Table  VIIIS). However, phenylacetyltransferase activity was stimulated by added potassium ions to a maximum observed at 0.1 M KCl; several other monovalent cations at 0.1 M also stimulated phenylacetyltransferase activity (Table VIE?). Both transferases had broad pH optima in the range of 8.4 to 8.6. Kinetics-The results of preliminary bisubstrate and product inhibition kinetic studies are summarized in Table II. Both transferases showed similar kinetic patterns. Double reciprocal plots of initial velocities uersus varying concentrations of either acyl-CoA or glycine at fixed concentrations of the other substrate revealed converging straight lines with a common intercept located to the left of the ordinate below or at the abscissa (Fig. 5s). This location of the intercept below or at the abscissa is consistent with a sequential reaction mechanism and the converging lines eliminate a classical pingpong mechanism where parallel lines would be expected (7). Double reciprocal product inhibition plots reveal that the peptide product acted kinetically as a competitive inhibitor with respect to the acyl-CoA substrate but a noncompetitive inhibitor with respect to glycine; the reduced CoA product produced a noncompetitive pattern with either the acyl-CoA or glycine as variable substrates (Figs. 6S and 7s). This pattern of inhibition is consistent with an ordered mechanism wherein the acyl-CoA substrate adds first and the peptide product dissociates last (7). DISCUSSION The study of acyl-CoA:amino acid N-acyltransferases in liver or kidney mitochondria appears important for several reasons. First, it may provide an explanation at the molecular level for the wide species diversity found in the excretion of amino acid conjugates of organic acids (1). Second, it should yield insights into the evolution and genetics of these important detoxifying systems. Third, it may relate to the pathogenesis and possibly the treatment of certain organic acidemias. Finally, it does provide a useful model for studying mechanisms of enzymatic transfer reactions.
The major contribution of the present study is the demonstration with nearly homogeneous preparations that the benzoyl-CoA:glycine N-acyltransferase and the phenylacetyl-CoA:glycine N-acyltransferase activities reflect the presence of two distinct enzymes in bovine liver mitochondria.
Both activities previously had been ascribed to a single protein (5).
Characterization of these two closely related transferases revealed several properties that were either unknown or different from those described before. For example, enzymatic acyl transfer to glycine was not completely specific as previously reported (5,8,9). Both L-asparagine and L-glutamine served as weak acyl acceptors posing the question as to whether or not the L-glutamine acceptor activity might account for the traces of phenylacetylglutamine found in cow's milk (10). Inhibition of bovine benzoyl-or phenylacetyltransferase activity by high unphysiological concentrations of divalent metal ions has not been reported nor have two substrate and product inhibition kinetic studies been performed. The differential sensitivity of these two enzymes to sulfhydryl reagents is a new finding as is the selective stimulation of phenylacetyltransferase activity by certain monovalent cations.
Our findings complement and generally substantiate those of Lau et al. (11) who recently studied the photoaffinity labeling of the bovine liver benzoyltransferase with p-azidobenzoyl-CoA.
These investigators obtained molecular weights of 36,000 by gel filtration on Sephadex G-100 and 34,000 by SDS-disc gel electrophoresis for a highly purified but nonhomogeneous preparation of this mitochondrial enzyme. Their value compares favorably with our previous estimate of 32,000 (12) and the present finding of -33,000 (Table 111s) Comparisons between corresponding acyltransferases of ox uersus rhesus monkey or human liver mitochondria reveal similarities and differences which relate to the evolution of these proteins. Specificities for acyl-CoA substrates and inhibitors were similar between species suggesting, not only that further enzyme purification will not change the acyl-CoA specificity reported for the primate enzymes (6), but also that the pattern of acyl-CoA specificity was established before the evolutionary diversion of primates from ungulates and possibly other mammals. In contrast, the interspecies differences in transferase specificities for amino acids reveal that the high specificity of the rhesus monkey and human benzoyl-and phenylacetyltransferases for glycine and L-glutamine, respectively, occurred after primate evolution diverged from that of the ungulates. The finding that purified bovine phenylacetyltransferase can utilize both glycine or L-glutamine as substrates leads to the speculation that a single mitochondrial transferase in New World primates may account for phenylacetyl conjugation to both glycine and L-glutamine. In contrast to Old World primates and man which excrete benzoates conjugated only with glycine and arylacetates conjugated only with L-glutamine, New World monkeys excrete both phenylacetylglycine and phenylacetylglutamine in the urine (13). The difference between the molecular weight of -33,000 found for the two bovine transferases compared to -24,000 for the rhesus monkey transferases (6) eliminates a single amino acid substitution and a monomer-dimer relationship as the basis for the interspecies dissimilarity between these enzymes. The small size of all of these proteins makes them attractive candidates for amino acid sequencing to study further the genetics and evolution of amino acid conjugation.
Acyltransfer from certain CoA esters could play a role in the pathogenesis of specific human organic acidemias. Bartlett and Gompertz have recently reported that CoA ester derivatives of certain aliphatic carboxylic acids implicated in these acidemias may serve as substrates for partially purified preparations of benzoyltransferase from bovine liver mitochondria (14). The present study shows in addition that activity with these nonpreferred substrates is attributable to the benzoyltransferase rather than the phenylacetyltransferase which undoubtedly contaminated their preparation. Although activation to the acyl-CoA derivative is the first and probably rate-limiting step in the pathway of amino acid conjugation (la), it is conceivable that the transfer step might become rate-limiting in certain genetic disorders. The preliminary two substrate and product inhibition kinetic studies are consistent with a sequential reaction mechanism wherein the acyl-CoA associates first with the enzyme, glycine adds to the enzyme prior to dissociation of the first product, CoASH, and the peptide product dissociates last. The kinetic data should, however, be interpreted with caution because insensitivity of current methods for product detection prevented our conducting experiments at acyl-CoA substrate concentrations well below their apparent Km values. Intersecting lines on the primary double reciprocal plots are inconsistent with a double displacement or ping-pong mechanism (7) and this mechanism involving an acyl-enzyme intermediate also seems unlikely because Lau et al. reported that photoafhnity-labeled benzoyltransferase contained both the arylacyl and the CoA moieties of p-azidobenzoyl-CoA (11). Lau's study also provides independent evidence for formation of a binary complex between the acyl-CoA and enzyme. This concept is supported by our earlier observation that benzoyl-CoA but not glycine protected benzoyltransferase from heat inactivation (15) and the present finding that phenylacetyl-CoA but not glycine protected phenylacetyltransferase from DTNB inhibition.
The kinetic studies suggest that peptide products should protect these enzymes too. Apparent Km values for substrates and products must eventually be compared with Ku values obtained directly from enzyme inactivation or spectroscopic studies. The marked differences in reaction rates of benzoyltransferase with different acyl-CoA substrates at near their kinetically saturating concentrations (Table VIS) argues against dissociation of CoA from the enzyme as being a final common rate-limiting step but detailed comparative kinetic studies with nonpreferred acyl-CoA substrates, e.g. butyryl-CoA, should be helpful in establishing this point.
The present purification procedure, perhaps combined with affinity chromatography (ll), permits the preparation of substrate quantities of at least the benzoyltransferase so that formation of binary complexes or putative enzyme-bound intermediates might be demonstrated directly or indirectly by appropriate spectroscopic techniques. Nuclear magnetic resonance studies of reduced coenzyme A as well as the acyl-CoA substrates and acyl-CoA inhibitors of both transferases have already been performed to facilitate these alternative approaches to elucidating the reaction mechanism (16,17).