Carbamylation of Aspartate Transaminase and the pK Value of the Active Site Lysyl Residue*

Abnormal lysyl residues can be detected in aspartate transaminase by following the rate of reaction of amino groups with KN’%O and the rate of enzymatic inactivation. Peptide isolation subsequent to carbamylation of the apoenzyme produces a peptide which is absent in the carbamylated holoenzyme. The composition of the carbamylated peptide matches that of a tryptic peptide containing the active site Lys-258. The holoenzyme retains full catalytic activity after carbamylation of its NHz-terminal alanine and lysyl residues other than Lys-258, which is protected by aldimine formation with pyridoxal phosphate. Apoenzyme prepared from KNCO-treated holoenzyme (apoenzyme’) is susceptible to further carbamylation at Lys-258 with irreversible loss of catalytic activity. Carbamylation of the active site lysyl residue is 25 to 50 times more rapid than that of the other 18 lysyl residues of aspartate transaminase. The kinetics of inactivation by KNCO at different pH values served to determine the pH-independent second order rate constant (k) and the pK of the amino group of Lys-258. These values are pK = 7.98 & 0.08 and k = 146 * 5 M-k’, which are similar to the values determined for carbamylation of the NH,-terminal groups of human hemoglobin R.

The determination of the pK values of amino acid residues at the active center of enzymes in solutiorl is a complex task. The methodology is largely limited to spectroscopic probing for the determination of an individual pK for those few cases where a distinctive spectroscopic property can be detected in the amino acid residue under investigation (l-3) or in kinetic procedures which follow the pK dependence of the reaction rate of a chemical modification for a specific residue in the enzyme (4)(5)(6)(7)(8)(9).
Lysyl residues in peptides show average pK values of -10. Several proteins, however, have shown unusually reactive lysyl residues with anomalous pK values which can be as low as 5.9 for oxalacetate decarboxylase (4) and 7.70 * 0.3 in glutamate dehydrogenase (7).
The unusual reactivity of some lysyl residues in proteins may manifest itself in their extraordinary reactivity with pyri-* This work was supported by Research Grant GM-20727 from the National Institutes of Health and the Indiana Heart Association.
$ Recipient of a Research Career Development Award from the National Institute of General Medical Sciences. doxal-P' with which they form Schiff bases that can be easily reduced and radioactively labeled (9)(10)(11)(12). This use of pyridoxal-P as a marker of specific regions of a protein in solution or of membranes is increasingly gaining acceptance as a tool for specific tagging of proteins. On the other hand, it now appears that pyridoxal-P may not be a specific reagent for lysyl residues only, since NH*-terminal groups in some proteins may also be susceptible to Schiff base formation with pyridoxal-P (13).
The family of the pyridoxal-P-dependent enzymes contains an unusually reactive lysyl residue which, with the exception of that in glycogen phosphorylase, is bound to the active site chromophore, pyridoxal-P, as an internal aldimine (14). The determination of the pK value of the unusual amine, which may be a prototype for the behavior of other less discriminating lysyl residues, has remained elusive. Only recently we have detected a group in the pyridoxal-P-binding region of aspartate 5664 The pK of Lys-258 in Aspartate Transaminase transaminase which ionizes with a pK value of 8.3 k 0.1 (15 fluid was obtained by mixing 1 liter of toluene (scintillation grade), 8 g of 2.5-diphenyloxazol, 0.5 g of 1,4-bis[2-(4-methyl-5-phenyloxazolyl)]benzene (scintillation grade), and 500 ml of Triton X-100; vials were routinely counted for 2 min. The paper chromatogram radioactivity was measured on a Packard model 7201 radiochromatogram scanner.

Inactivation of Pig Heart Supernatant
Aspartate Transaminase by Cyanate-The susceptibility of the aspartate transaminase to inactivation by reaction with 0.2 M KNCO depends on the presence or absence of co-enzyme. Holoenzyme with the active site lysyl residue blocked by pyridoxal-P, as an internal aldimine, is functionally unaffected by the cyanate treatment. On the other hand, the same cyante-treated holoenzyme, after conversion to its apoenzyme form (apoenzyme') can be subjected to the same cyanate treatment with a resulting rapid and total inactivation with a pseudo-first order rate constant of 0.09 min-' at 37" (Fig. 1). The same rate of inactivation is observed on cyanate treatment of apoenzyme produced from native, untreated holoenzyme. No recovery of activity can be detected by extensive dialysis or passage through a Sephadex G-25 column for either apoenzyme. Incubation of the cyanatetreated apoenzymes with 1 x lo-' M pyridoxal-P is also ineffective in restoring catalytic activity.
The rates of inactivation by cyanate treatment, at constant pH and enzyme concentration, are dependent on the cyanate concentrations.
Plots of log of activity uersus time of cyanate treatment are linear, indicating their first order dependence. The pseudo-first order rate constant observed (k,,.) follows a linear relationship with cyanate concentration characteristic of an overall second order reaction (Fig. 2). Under the same experimental conditions no activity losses are detected when either apoenzyme is incubated in the absence of KNCG.

Incorporation of ['"C]Cyanate
to Apo-and Holoenzyme-The pattern of incorporation of ['"Clcyanate into transaminase depends on the form of enzyme used (Fig. 3). In all cases, holo-and apoenzymes, we can distinguish at least two phases representing groups of residues with diverse susceptibility to cyanate. At pH 7.4 most of the fast-reacting groups react within the first 20 min of treatment with 0.2 M KN"C0. In the apoenzyme the number of groups modified in the initial phase is less than 3/subunit. In the holoenzyme the fast phase corresponds to incorporation into 1 residue. However, in apoenzyme', produced from extensively cyanate-treated holoenzyme, the same biphasic behavior is observed (Line 2, Fig. 3   ' Conditions as in Figs. 1 and 3. b "C-labeled compound. P 6 6( the fast phase. The cyanate is always incorporated by carbamylation of amino groups. The groups modified are the NH*-terminal alanine residue, identified after preparation and hydrolysis of its hydantoin derivative, and the t-amino groups of lysyl residues (Table I). After KN'"C0 treatment of apoenzyme and holoenzyme the carbamylated NHz+erminai residue contains radioactivity but it is absent in "C-carbamylated apoenzyme*. None of the radioactivity is removed by dialysis or passage through the Sephadex G-25 column and the values of "C are identical whether measured after trichloroacetic acid precipitation of the protein or passage through the Sephadex column.
The initial phase of holoenzyme cyanate incorporation represents carbamylation of the NHz-terminal alanine. The rest of the modification can be accounted for by the appearance of radioactive homocitrulline in the chromatography of acid hydrolysates (Fig. 4). The same amino acids are modified during '"C carbamylation of apoenzyme except for the total greater "C incorporation during the fast phase which is to be expected if unusual lysyl groups become available after removal of pyridoxal-P.
In apoenzyme* the only amino acid radioactively modified appears to be lysine detected as 14Clabeled homocitrulline.
The differences between the fast, and slow rates of carbamylation of the two types of lysyl residues in apoenzyme are between 20 and 50 times if we assume (see below) that the fast rate corresponds to carbamylation of the active site lysine, Lys-258, and the slow phase to random carbamylation of the other 18 lysyl residues in each subunit.
The fast rate of cyanate incorporation, after correction for the slow contribution, is k -0.12 min-' which is close to the first order rate constant of inactivation, k = 0.09 min-', obtained under identical experimental conditions.

Kinetics of pH Dependence of Enzyme Inactivation by
Cyanate-The inactivation of aspartate transaminase by cyanate follows a pseudo-first order behavior at the pH tested between pH 7.59 and 8.91 (Fig. 5). A plot of these rate constants uersus pH is sigmoidal with an inflexion point at pH 8  Peptide-The identification of the amino acid residues modified by cyanate was accomplished by isolation and characterization of the 14Clabeled peptide. Prior to tryptic hydrolysis both apoenzyme and apoenzyme* were carboxymethylated as indicated under "Experimental Procedure." Passage of the tryptic peptides through a Sephadex G-50 column produced the elution profile in Fig. 6. The fractions containing the main radioactivity peak were pooled to be purified in the DEAE-cellulose column, which gave the elution profile in Fig. 7. The radioactive material appears mainly at the beginning of the first gradient and in a later fraction at high salt concentration.
In the apoenzyme most of the radioactivity belongs to the first fraction. On the other hand, previous treatment of the holoenzyme with nonradioactive KNCO produces apoenzyme, which after treatment with ["Clcyanate shows an elution pattern as at the bottom of Fig. 7, with a marked decrease of radioactivity in the first peak. The differences in radioactivity of the first peak are also consistent with the total diminution of radioactivity incorporated (Table I and Fig. 3) by apoenzyme*. The fractions between tubes 80 and 90 contain peptides with identical amounts of radioactivity in apoenzyme or apoenzyme*. Amino acid composition of this fraction did not match any tryptic peptide in the sequence of aspartate transaminase (20, 21). Further purification in a Sephadex G-10 column and paper chromatography shows a single radioactive fluorescamine and/or ninhydrin-positive spot which upon acid hydrol- Top, apoenzyme; Bot-Absorbance; 04, radioactivity. tom, apoenzyme*.
by guest on July 8, 2020 http://www.jbc.org/ Downloaded from ysis had the composition (Table II) of the tryptic peptide containing Lys-258, which in native transaminase has the following sequence: Try-Phe-Val-Ser-Glu-Gly-Phe-Glu-Leu-Phe-Cys-Ala-Gln-Ser-Phe-Ser-Phe-Ser-Lys-Asn-Phe-Gly-Leu-Try-Asn-Glu-Arg. Tryptic hydrolysates of KN14CO-treated holoenzyme produce a DEAE-cellulose elution profile where Fractions 80 to 90 are void of radioactivity and where the large early radioactivity peak is present.
Binding of the co-enzyme can be determined by fluorescence quenching (15,22). By this method we detect binding of one pyridoxal-P per subunit of transaminase. Conversion of Lys-258 to homocitrulline blocks the amine which is thus unable to form a Schiff base with the aldehyde of pyridoxal-P, hence, the unusual spectrum with a maximum of 390 nm (Fig. 8). The union with the co-enzyme is strong and separation cannot be achieved by dialysis or passage through a Sephadex G-25 column. Conversion of the amine to the ureido group does not appear to create much steric hindrance at the (1) (1) a Represent the numbers of residues relative to valine. active site. The modified enzyme binds co-enzyme, and addition of an amino acid substrate produces the immediate formation of a complex with the characteristic absorption maximum (430 nm) of enzyme.substrate complexes (23) where the substrate amino acid forms an internal Schiff base with the co-enzyme. The complex does not appear to be of an abortive type since it is able to convert, albeit very slowly, to a pyridoxamine complex spectrum with absorbance at 330 nm. Monitoring the properties of the latter mixture shows production of keto acid.' DISCUSSION Cyanate treatment of proteins is most likely to result in stable derivatives of amino groups such as the c-amino group, of lysine or the terminal u-NH2 group (19,24). The products of cyanate reaction with other amino acid residues decompose in aqueous solutions at neutral pH (19,24).
In supernatant aspartate transaminase KNCO reacts with both NH*-terminal and lysyl residues with formation of carbamylated alanine (NH,-terminal residue) and homocitrulline. Selection of the groups modified is based on the dissimilarity of pK values of different amino groups, since carbamylation requires the nonprotonated amine (19,24). Susceptibility to carbamylation is further enhanced since at least one lysine is protected by the pyridoxal-P chromophore. Thus, carbamylation of holoenzyme should lead, as observed, to modification of the NHz-terminal alanine and some of the nonactive site lysyl groups. The active site Lys-258 becomes exposed only in the apoenzyme and if its pK is abnormal preferential carbamylation of Lys-258 should occur.
The pH dependence of inactivation of the apoenzyme* is consistent with carbamylation at amino groups where isocyanate reacts with an unprotonated amine with a pK value of 7.98 + 0.08 (Fig. 5). The data are most consistent with assigning this pK to the e-amino group of Lys-258. The pK value as determined by the chemical modification method is similar to that value detected for a group at the active site using an NMR spectroscopic procedure after the insertion of a 19F marked probe (15). The pH independent second order rate constant, k = 148 f 5 Mm%', calculated from the data in Fig. 5 is within the range that can be expected for cyanate reactivity with amines of similar pK (7,8 by guest on July 8, 2020 http://www.jbc.org/ Downloaded from steric hindrance or spectrophotometric anomality in the active site chromophore is observed. The method of using the pH dependence of the rate of carbamylation appears to be a useful procedure in detecting unusual lysyl residues with pK values considerably lower than those in the rest of the protein (7,8).
Since the kinetics of inactivation give no evidence of deviation from pseudo-first order kinetics, the possibility of hindrances or of any cooperativity effects by the other residues reacting with cyanate is unlikely. Conformational or other structural changes induced by the modification of the nonactive site lysine residue or the NHz-terminal group are ruled out by the retention of catalytic activity in carbamylated holoenzyme. These observations are consistent with other results in which acetylation or succinylation of almost 80% of all amino groups in aspartate transaminase does not significantly alter catalytic activity (26)(27)(28). In the apoenzyme the effect of some carbamylatian of amino groups extraneous to the active site is more difficult to assess. However, since after cyanate treatment pyridoxal-P binds stoichiometrically and the resulting holoenzymes appear to undergo a half transamination upon addition of amino acid, a large conformational change is unlikely.
All enzymes in which pyridoxal-P participates in catalysis bind the co-enzyme as the Schiff base with the c-amino group of a lysyl residue. Among these enzymes the transaminases not only bind pyridoxal-P but also pyridoxamine-P.
From these facts and the present observations it can be concluded that the active site lysine is not a major contributor to the equilibrium binding constant in transaminases. The lysyl residue must also occupy a spot with a considerable degree of freedom in the active site topography, where not even the introduction of the carbamyl moiety significantly perturbs pyridoxal-P. After carbamylation of Lys-258 the co-enzyme seems to form the internal aldimine with an amino acid substrate with full retention of initial steric discrimination, as D amino acids are not accepted.' The initial enzyme.co-enzyme.substrate complex of holoenzyme carbamylated at Lys-258 must be held in a mostly productive configuration in the active center cage as the complex is still able to slowly transaminate. The creation of a new form of the enzyme (active site carbamylated) with a co-enzyme attached at the active site and able to accept substrate may yet provide the best tool for the study of the functional role of Lys-258.
Finally, we should note that this new awareness of low pK t-amino groups in pyridoxal-P proteins should be considered as a possible trouble spot in cyanate treatment of sickle cell patients (29). The susceptibility of the unusual lysyl residues to carbamylation is of the same order of magnitude as reported for the NH*-terminal group of human hemoglobin. The necessary long term cyanate therapy in sickle cell patients is bound to perturb the pyridoxal-P enzymes, particularly those enzymes with low affinity for pyridoxal-P or in individuals with lbw pyridoxal diets or pathologic conditions where a protective effect of the co-enzyme is likely to be minimal. In view of the reports that some neurological effects are observed in trial patients (30) subjected to cyanate treatment and a significant fraction of administered cyanate appears in the organs of experimental animals (29), and in view of the fact that pyridoxal-P plays an essential role in the metabolism of biogenic amines, the data reported in this paper should be explored further. Administration of large doses of Vitamin B, to cyanate-treated sickle cell patients appears to be a worthwhile meliorative procedure.