Alkyl isocyanates as active site-directed inactivators of guinea pig liver transglutaminase.

Alkyl isocyanates are effective inactivators of guinea pig liver transglutaminase. Based on the specificity of the reaction the protection against inactivation by glutamine substrate, and the essential nature of calcium for the inactivation reaction, it is concluded that these reagents act as amide substrate analogs and, thus function in an active site-specific manner. Support for the contention that inactivation results from alkyl thiocarbamate ester formation through the single active site sulfhydryl group of the enzyme is (a) the loss of one free--SH group and the incorporation of 1 mol of reagent/mol of enzyme in the reaction, (b) similarity in chemical properties of the inactive enzyme derivative formed to those previously reported for another alkyl thiocarbamoylenzyme and an alkyl thiocarbamoylcysteine derivative, and (c) the finding that labeled peptides from digests of [methyl-14C]thiocarbamoyltransglutaminase and those from digests of iodoacetamide-inactivated enzyme occupy similar positions on peptide maps. Transglutaminase was found to be inactivated neither by urethan anlogs of its active ester substrates nor by urea analogs of its amide substrates. It is concluded on the basis of these findings that inactive carbamoylenzyme derivatives are formed only by direct addition of the transglutaminase active--SH group to the isocyanate C--N double bond, and not, like several serine active site enzymes, by nucleophilic displacement with urethan analogs of substrate, or by nucleophilic displacement with urea analogs of substrate.

The transglutaminases catalyze hydrolysis and aminolysis of the carboxamide group of peptide-bound glutamine residues (for review, see Ref. 2). The enzymes also catalyze these reactions with active esters, e.g. p-nitrophenyl esters (3,4) and certain thioesters (5). In addition, some esters of aliphatic alcohols which are not active esters have been found to function as substrates for the guinea pig liver enzyme (6). Studies in this laboratory have provided evidence that all of these reactions proceed through acyl-enzyme intermediates formed between the essential -SH group in the Ca*+-activated enzymes and the acyl portions of substrates (for review, see Ref. 2). The kinetics of hydrolysis and transfer conform to a modified double displacement mechanism in which acylenzyme is partitioned between water and another nucleophile, e.g. a primary amine.
Recent investigations on the substrate specificity of guinea pig liver transglutaminase have led to the following active site model for this enzyme (7)(8)(9). Polypeptides containing a substrate glutamine residue bind in a single direction along the * This communication is the 10th paper in the series, "Mechanism of Action of Guinea Pig Liver Transglutaminase." The preceding paper is Ref. 6. This work was presented in part at the Biochemistry/Biophysics Meeting, Minneapolis, Minnesota, June 2 to 7, 1974 (1). $ Present address: Division of Research Grants, National Institutes of Health, Bethesda, Md. 20014. enzyme surface; the side chains of amino acids over a range of residues on each side of the substrate glutamine residue exert an influence on catalysis. The arrangement of the substrate peptide chain on the enzyme surface directs the specificity of the enzyme for glutamine residues of the L configuration; the a-hydrogen of an L-glutamine residue is on the side of this substrate that is directed away from the enzyme surface, whereas the a-hydrogen of a n-glutamine residue abuts the enzyme surface resulting in misalignment of the carboxamide side chain in its binding pocket and, thus, in a concomitant loss in catalytic efficiency. The hydrophobic binding pocket for glutamine side chain assumes dimensions of approximately 5 x 5 A at some stage of catalysis. At the apex of this pocket is located the enzyme sulfhydryl group which participates in acyl-enzyme intermediate formation. The use of alkyl isocyanates as active site-specific inactivators for serine proteases (10) and for the -SH enzyme, yeast alcohol dehydrogenase (ll), has been introduced by Wold and co-workers. They have provided evidence that, following binding of these substrate analogs, inactivation occurs as a result of formation of stable N-alkyl urethans with the hydroxyl group of the active site serine in chymotrypsin and elastase (12). In the case of the dehydrogenase, analogous N-alkyl thiourethans result from reaction of isocyanate with three enzyme sulfhydryl groups (11). recording spectrophotometer. Radioactivity was measured by standard methods in Hydromix counting fluid (Yorktown Research) with a Beckman LS-233 liquid scintillation spectrometer.

Inactivation
of Transglutaminase by Methyl Isocyanate - Fig.  1 is a typical inactivation pattern obtained upon treatment of the enzyme with increasing levels of methyl isocyanate in the presence of Ca2+ at pH 5.7. It is evident from these data that loss in enzymatic activity is a stoichiometric function of the amount of isocyanate used and that total inactivation occurs as a result of addition of 1 mol of reagent/ mol of enzyme. The assays of Fig. 1 were performed 2 min after addition of reagent. Longer incubation with reagent (up to 30 min) caused no further loss in activity. The very rapid nature of the inactivation reaction was evidenced from the fact that, at 1 eq of reagent to enzyme, total loss in activity, as measured by the continuous spectral esterase method, was observed within the time necessary to commence the assay (20 to 30 s).
Essentially, the same stoichiometry and rapid rate of inactivation with methyl isocyanate were observed at pH 7.0. At pH 7.5 and above, higher levels of reagent were required for inactivation. This is probably due to the rapid loss of reagent by hydrolysis, the rate of which may become comparable to, or exceed, that of enzyme inactivation.
Brown and Wold (10)   in the absence of Ca*+ was observed only after addition of 6 to 8 eq of reagents.
Under the conditions of Fig. 1, the transglutaminase substrate, Z-L-glutaminylglycine showed a pronounced protection against inactivation.
At the level of 1 eq of isocyanate, addition of 20 and 40 mM substrate reduced the per cent inactivation to 78 and 52, respectively.
Specific Nature of Isocyanate Inactivation-In addition to methyl isocyanate, several other aliphatic isocyanates were found to inactivate transglutaminase (see later section, "Inactivation by Various Aliphatic Isocyanates").
The report of Twu and Wold (11) that the inactivation of yeast alcohol dehydrogenase by n-butyl isocyanate is a consequence of thiocarbamate ester formation through essential -SH groups of the enzyme suggests that isocyanate inactivation of transglutaminase may also occur by way of a thiocarbamylation reaction. If so, however, this must involve only a single essential enzyme --SH group. This contention is based on the present observation that transglutaminase is fully inactivated by treatment with one equivalent of methyl isocyanate (Fig. 1). It is given support by the earlier findings that alkylation, e.g. by iodoacetamide (14,19), or trimethylacetylation (20) of a single essential -SH group in the enzyme leads to complete inactivation. Two samples of transglutaminase that had been fully inactivated by treatment with equimolar amounts of methyl isocyanate as outlined in Fig. 1 were examined for total -SH content by the use of the spectrophotometric dithiodipyridine procedure. Values of 17.0 and 17.4 -SH groups/molecule were found. Comparison with values of 18.3 and 18.7 -SH groups/ molecule, obtained by the same procedure for samples of the untreated enzyme, supplies strong evidence that inactivation by methyl isocyanate results from reaction of this reagent with an -SH group in the enzyme.
Reaction of isocyanates with sulfhydryl compounds occurs by nucleophilic addition to the C-N double bond (Reaction 1). The thiocarbamoyl enzymes that would be formed by reaction of isocyanates with the single essential -SH group HO I II R-SH + R'-N=C=O-+ R'-N-C-S-R (1) of transglutaminase bear structural resemblance to the thiolester acyl-enzyme intermediates believed to occur during the course of normal catalysis (see introduction).
It seems possible that inactive thiocarbamoyl enzyme derivatives could be formed by nucleophilic displacement reactions with monosubstituted urea analogs or urethan analogs of transglutaminase substrates according to Reactions 2 and 3, respectively.

HO HO
R-SH + R&-NH* -R&-S-R + NH3 (2) The very rapid rate and stoichiometry of enzyme inactivation by methyl isocyanate and the long half-life of this reagent in water at pH 5.7 precludes N,N'-dimethyl urea as an enzyme inactivating agent. This disubstituted urea could arise via Reactions 4a and 4b.
CHsN=C=O + HOH-+CH3NHp + CO;, CHsN=C=O + CH3NH2-CH3-NH$NH-CH3 (4b) 0 7696 N-Methylurea at levels between 1 and 10,000 mol/mol of enzyme was found to cause no enzyme inactivation at pH 5.7 or 7.0 (buffer conditions of Fig. 1). This urea is the structural analog of n-propionamide, a substrate for transglutaminase (2). Because the affinity of the enzyme for aliphatic amides is low and since that for the urea may be even lower, it was judged worthwhile to test the urea analog of the glutamine substrate, Z-L-glutaminylglycine.
This analog, Z-Lalbizzinylglycine did not inactivate the enzyme nor did it function as a substrate. That the albizzin derivative was bound to the active site of the enzyme is evident from the finding that it showed linear competitive inhibition against Z-L-glutaminylglycine (Fig. 2). The inhibitor constant, KL,, estimated from the data of Fig. 2, was 67 f 5 mM, a value similar to the apparent K, of 66 f 4 mM for the substrate under these experimental conditions.
The urethan, p-nitrophenyl N-methylcarbamate, is a structural analog of p-nitrophenyl propionate, one of the active ester substrates for transglutaminase (5). Incubation of equimolar amounts of the enzyme and this urethan at pH 5.7 and 25" in the presence of Ca*+ (other conditions of Fig. 1) caused no loss in enzyme activity over a period of 2 hours. When the pH was raised to 7.0, however, a slow loss in enzyme activity and concomitant release of p-nitrophenol (as measured by absorbance at 400 nm) was observed. The release of p-nitrophenol was complete in approximately 60 min, at which time only 2 to 5% of the enzyme activity remained. In control experiments under the same conditions, except without enzyme, the rate of formation of p-nitrophenol was found to be exactly the same as in the inactivation experiment. This suggests that inactivation of the enzyme does not occur by direct reaction between the urethan and the enzyme, but rather, that a nonenzymatic conversion product of the urethan is the true inactivator. Bender and Homer (16)  On the basis of kinetic studies, Christenson (21) suggested that alkaline hydrolysis of NJ-dialkyl urethans proceeds by a very different mechanism (Reaction 6) than that shown above for the N-monoalkyl urethan (Reaction 5).

NJ-dimethylcarbamate
is the dialkylcarbamate analog of p-nitrophenyl isobutyrate, an active ester substrate for transglutaminase (5). This urethan was found to cause no inactivation of transglutaminase, nor was any p-nitrophenol formed over a period of 2 hours at pH 7 and 25" in the presence of Ca*+ at molar ratios of urethan to enzyme between 1 and 10.
Stoichiometry and Position of "C Incorporation by "Clabeled Methylisocyanute, Reversal of Inactivation-Because of the volatile nature of methyl isocyanate (b.p. 59.6"), no attempt was made to prepare this compound in a radioactively labeled form. Instead, p-nitrophenyl N-["Clmethylcarbamate, which could be easily prepared and stored, was used as a source of the labeled isocyanate. It is reasonably assumed on the basis of the foregoing data that it is indeed methyl isocyanate, formed by alkaline breakdown of this urethan, that is the enzyme inactivating agent. Two samples of enzyme were treated at pH 7.0 and 25" in the presence of Ca2+ with 1 eq of "C-labeled p-nitrophenyl N-methylcarbamate.
After 1 hour the samples were gel-filtered through a column of Sephadex G-25 in 0.05 M Tris-acetate buffer, pH 7.0, and portions of the protein fractions were analyzed for protein concentration, radioactivity, enzymatic activity, and -SH content. The analysis showed 0.95 and 1.02 mol of reagent incorporated/mol of enzyme, 4 and 0% of the initial enzymatic activity, and 17.4 and 17.5 -SH groups/ molecule, respectively.
The untreated enzyme samples were found to contain 18.3 and 18.5 -SH groups/molecule, respectively.
The solution of inactivated enzyme containing 0.95 mol of "C-labeled reagent/m01 was concentrated by ultrafiltration (UM-10 Diaflo filter, Amicon) and adjusted to pH 8.5. After esters. The hydrolysis of all aliphatic carbamate esters apparently -..
-'It has been pointed out (16) that this dichotomy holds only for aryl follows ReactIon ti. 7697 incubation for 18 hours at 25", the solution was gel filtered as outlined above and portions of the protein peak were analyzed for protein concentration, radioactivity, and enzymatic activity; the salt fraction was tested for radioactivity.
The protein was found to contain only 30% of the initial radioactivity (0.29 mol of "C-labeled reagent/mol); the remainder of the radioactivity was accounted for in the salt fraction. The enzymatic activity was partially restored, showing 35% of that of the unmodified enzyme. Peptide maps were prepared as outlined previously (22) from enzymatic digests of a sample of the ["Clmethyl isocyanatelabeled enzyme and from a sample of enzyme that was fully inactivated by the incorporation of 1 eq of 'C-labeled iodoacetamide (14,19). Digestion of these labeled enzyme samples was carried out as outlined earlier (lS), except that approximately 1: 1 molar ratios of digestive enzymes to labeled substrates were used and digestion times were limited to 1 hour for trypsin followed by 2 hours for chymotrypsin.
The shorter digestion times were employed in order to minimize losses in radioactivity from the isocyanate-labeled enzyme. Autoradiograms of the peptide maps prepared from each of the labeled enzyme samples showed two areas of radioactivity in positions identical to those observed previously for 16 to 20-hour trypsin-chymotrypsin digests of the iodoacetamide-inactivated enzyme (R, values of 0.53 and 0.65 in 1-butanol, acetic acid, water and in the area of the neutral amino acids in pH 3.5 electrophoresis) (23). Each of the areas in both peptide maps gave positive reactions for tryptophan residues by the use of p-dimethylaminobenzaldehyde in HCl. An additional radioactive area (R, approximately 0.9; in the position of neutral amino acids in electrophoresis) was observed on the peptide map from the isocyanate-inactivated enzyme. This area showed no reaction for tryptophan.
The two peptides from iodoacetamide-inactivated enzyme have been identified as Gly-Gln-Cys(S-carbamidomethyl)-Trp and Tyr-Gly-Gln-Cys(S-carbamidomethyl)-Trp (19). It is likely that two of those found in the digest of isocyanate inactivated enzyme are the N-methylcarbamoyl counterparts.
The carbamidomethyl and methylcarbamoyl group are structurally similar. One would expect S-carbamidomethyl and N-methylcarbamoyl peptides of the same amino acid sequence to occupy the same or very similar positions on peptide maps. The additional radioactive material found with the isocyanate-inactivated enzyme was not identified. However, it occupied a position on the peptide map very different from that of methylamine which would be expected as a product of alkaline hydrolysis of the enzyme methylthiocarbamate group.

Inactivation by Various Aliphatic
Isocyanates-It is apparent from experiments reported above that the rate of inactivation of transglutaminase by methyl isocyanate is too rapid to measure by the use of conventional sampling techniques and available enzyme assays. Each of the alkyl isocyanates tested, with the exception of isopropyl isocyanate, also was found to inactivate the enzyme very rapidly. Therefore, no attempt was made to determine rate constants for inactivation.
Under the conditions of Fig. 1 and at the level of 1 mol/mol of enzyme, ethyl isocyanate, propyl isocyanate, butyl isocyanate, isobutyl isocyanate, pentyl isocyanate, and isopentyl isocyanate each reduced the level of catalytic activity to less than 25% of the initial value within 3 min after addition. Isopropyl isocyanate, on the other hand, effected less than 10% inactivation after 10 min.
group of reduced glutathione was examined. The results are given in Fig. 3. Here the -SH and isocyanate concentrations are 40-fold higher than those in the enzyme inactivation mixtures. The slow rates of reaction of the isocyanates with the glutathione -SH group compared to the rapid inactivation of enzyme by all except isopropyl isocyanate are in agreement with active site-directed enzyme inactivation. Although the rate of reaction of isopropyl isocyanate with the glutathione -SH group is only about one-half to two-thirds that of the other isocyanates, this does not account for its very much lower degree of enzyme inactivation. DISCUSSION The findings reported here supply evidence that alkyl isocyanates inactivate guinea pig liver transglutaminase as a consequence of N-alkyl thiocarbamoyl ester formation with the single active site sulfhydryl group of the enzyme. The evidence for this active-site-specific reaction is briefly as follows. (a) Treatment of the enzyme in the presence of Ca2+ with 1 eq of methyl isocyanate results in rapid total loss in catalytic activity and an accompanying loss in one -SH group in the enzyme. With p-nitrophenyl ["C]methylcarbamate as a source of labeled methyl isocyanate, the incorporation of 1 mol of reagent/m01 of enzyme was observed. (b) The substrate, Z-L-glutaminylglycine, effectively protects against methyl isocyanate inactivation.
(c) As in the case of other reagents that inactivate transglutaminase by reaction with its active site -SH group, e.g. iodoacetamide (14,19) and a-bromo-4hydroxy-3-nitroacetophenone (24), isocyanate inactivation is less specific in the absence of the catalytically essential metal ion, Ca2+. (d) The loss of label and partial recovery of enzymatic activity that occurs upon alkaline treatment of the "C-labeled inactive enzyme is characteristic only of the N-alkylcarbamoylcysteine derivative (10). Butylcarbamoyl derivatives of lysine, tryosine, and serine have been shown to be completely stable to incubation at pH 10.5 and 37".' (e) The