/. Biochem., 81, 395-402 (1977)

The Reactions of Phenylglyoxal and Related Reagents with Amino Acids1 Kenji TAKAHASHI5 Department of Biophysics and Biochemistry, Faculty of Science, University of Tokyo, Bunkyo-ku, Tokyo 113 Received for publication, June 22, 1976

1. The reaction of phenylglyoxal (PGO), glyoxal (GO), and methylglyoxal (MGO) with amino acids were investigated at mild pH values at 25°. These aldehydes reacted most rapidly with arginine and the rate of reaction increased with increasing pH values. Histidine, cystine, glycine, tryptophan, asparagine, glutamine, and lysine reacted with these aldehydes at significant but various rates, depending on the pH and the kind of the reagent used. The reactions with these amino acids seemed to involve both the a-amino groups and the side chain groups, and no significant reaction appeared to occur with the side chain alone except with those of arginine, lysine, and cysteine. These reagents were similarly reactive with the guanidinium group of arginine, but PGO appeared to be much less reactive with the E-amino group of lysine than MGO and GO. The other ordinary amino acids were very much less reactive or did not react at all with these reagents, with the exception of cysteine. 2. Di-PGO-L-arginine was prepared from N"-benzyloxycarbonyl-L-arginine, and di-PGOmethylguanidine from methylguanidine, and the stoichiometry of the reaction of two PGO molecules with one guanidino group was confirmed. A glyoxal derivative of L-arginine (GOarginine) was prepared by reaction of glyoxal with arginine. GO-arginine was fairly unstable, especially at higher pH values. A similar derivative (MGO-arginine) was also found to be formed by reaction of MGO with L-arginine, and was similarly unstable. These derivatives, however, did not regenerate arginine upon acid hydrolysis.

In the previous study (1), phenylglyoxal was shown to be a useful reagent for specific modification of arginine residues in proteins. Glyoxal and methyl-

glyoxal were also shown to be useful for the same purpose (1,2). These aldehydes may react also with a-amino groups and certain side chain groups of amino acids other than arginine, depending on the reaction conditions and the kind of reagent * A preliminary account of this study appeared in the ^ In ^ c o n n e c t i o n > it seemed intere sting to abstract of the 22nd Symposium on Protein Structure, . , ., .. .. .. f .• . e X a m J e 1D SOmC d e t a Sendai, .971, pp. 81-84 (in Japanese). " '' the reactlvlties °f.these • Present address: Department of Biochemistry, Pn- r e a 8 e n t s W I t h a m j n 0 a C l d s u n d e r v a n o u s c o n d l t i o n s mate Research Institute, Kyoto University, Inuyama, S o m e o f t h e r e s u l t s o f t h i s examination were reAichi 484. ported in a previous paper (7). Abbreviations: PGO, phenylglyoxal; GO, glyoxal; This paper describes in detail the results obMGO, methylglyoxal. tained from the reactions of phenylglyoxal, glyoxal, Vol. 81, No. 2, 1977

395

396

K. TAKAHASHI

and methylglyoxal with amino acids, with special reference to their reactions with arginine. The results of further studies on the reaction of these reagents with proteins are described in a succeeding paper (3).

ditions as above, except that the final concentration of tryptophan was twice as high as those of the other amino acids. The reactions of GO and MGO with the amino acid mixture were performed under the same conditions unless otherwise specified. To examine the rates of reactions of GO and MGO with arginine EXPERIMENTAL and lysine separately, 400 p\ of a 1 % aq. solution Materials—Phenylglyoxal (PGO) hydrate was of arginine-HO or lysine-HQ was mixed with a purchased from K and K Laboratories, New York, 3 % solution of GO or MGO in 0.2 M buffer of the glyoxal (GO) hydrate from Mann Research La- desired pH. The amounts of ninhydrin-positive boratories, New York, and methylglyoxal (MGO) derivatives of arginine formed by reaction with GO (45% aq. solution) from Aldnch Chemical Co., and MGO were calculated using the ninhydrin New Jersey. Amino acid calibration mixture, type color value for arginine. In the case of GO, the I, was obtained from Beckman Instruments Inc., reactions with N-1-methylhistidine and N"-acetylCalifornia, L-arginine hydrochloride (A grade) histidine were also performed separately under from Calbiochem, California, and N°-benzyloxy- similar conditions at pH 8.2 and 7.0. carbonyl-L-arginine • HC1 and methylguanidinePreparation of Di-PGO-N '-benzyloxycarbonylHC1 from Mann Research Laboratories. L-arginine and Di-PGO-L-arginine-2HBr—N'-BenAmino Acid Analysis—Amino acids were de- zyloxycarbonyl-L-arginine-HCl, 69 mg (0.2mmol), termined by the automatic procedure of Spackman was taken into a test tube and dissolved in 0.2 ml et al. (4) with a Beckman-Spinco model MS amino of 1 M NaHCO,, and to this was added 0.5 ml of 0.2 M N-ethylmorphohne acetate buffer, pH 8.0, acid analyzer. Reactions of Phenylglyoxal and Related Re- and 0.7 ml of methanol. To this mixture was agents with Amino Acids and Their Derivatives— added dropwise, under magnetic stirring, 154 mg To determine the rates of reaction of PGO with (1 mmol) of PGO-hydrate dissolved in a mixture free amino acids, 200 pi\ of Beckman-Spinco ammo of 2.5 ml of the above buffer and 2.5 ml of methaacid calibration mixture, type 1 (2.5 fimol of each nol. After 6 h the reaction mixture became turbid. amino acid per ml), was diluted with 2.0 ml of each The reaction was allowed to proceed at room buffer solution and to the mixture was added 2.0 temperature for a total period of 20 h in the dark. ml of a 3 % solution of PGO hydrate, dissolved in The reaction mixture was then transferred to a the same buffer, containing 10% methanol.8 The round-bottomed flask and methanol was removed pH of the reagent solution was adjusted to that of by rotary evaporation. The remaining aq. soluthe buffer solution with dilute NaOH or acetic tion, containing a sticky precipitate, was shaken acid, if required. The buffers used were 0.05 M twice with 10-15 ml of ether and then rotary Na.COj-0.1 M NaHCO3, pH 9.0; 0.2 M NaHCO,, evaporated to dryness below 40°. The residue was pH 8.1 or 8.2; 0.2 M N-ethylmorphohne acetate, dissolved in 1 ml of methanol and transferred to a pH 8.0; 0.2 M sodium phosphate, pH 7.0; 0.2 M test tube. To this was added about 20 volumes of sodium maleate, pH 6 5; and 0.2 M sodium acetate, ether to get a sticky white precipitate. After pH 5.5. The reaction mixture was kept at 25° for removal of the ether, the precipitate was again 24 h in the dark. At intervals, aliquots of 500 p\ dissolved in 0.5 ml of methanol and to this were were pipetted into 2.5 ml of 0.2 M sodium citrate added about 10 volumes of ether to get a white buffer, pH 2.2, and the mixtures were stored frozen crystalline precipitate. The precipitate was washed at —10°. These samples were subsequently ana- in 4 ml of water and centrifuged. This washing lyzed on an amino acid analyzer. The reactions was repeated twice more and the precipitate was of PGO with asparagine, glutamine, and tryptophan then dried in an evacuated desiccator over CaCl, were performed separately but under the same con- at room temperature. The yield of di-PGO-N'benzyloxycarbonyl-L-arginine was 76 mg (66 %). Calculated for QoHwO,N4 • H,O (576.62) : C 62.49, ' No methanol was needed in the GO and MGO reac- H 5.59, N 9.72. Found: C 61.25, H 5.55, N 10.23. tions. J. Biochem.

397

PGO, GO, & MGO REACTIONS WITH AMINO ACIDS To a portion (33 mg) of the substance obtained above was added 0.4 ml of 30% HBr in glacial acetic acid in a centrifuge tube and the mixture was kept at room temperature for 1 h in a dessicator over CaCl,. The mixture was then chilled in ice water and ether was added m order to obtain a precipitate. The ether was removed by centrifugation. The slightly yellowish white precipitate was washed with 5-7 ml of ether four more tunes and then dried in an evacuated desiccator over CaG t at room temperature. The yield of di-PGO-Larginine-2HBr was 34 mg (98%). Calculated for CMHl4O6N1-H,O-2HBr (604.316):C 43.73, H4.67, N 9.27. Found: C 42.21, H 4.50, N 9.06. Preparation of Di-PGO-methylguanidine— Methylguanidine-HCl, 550 mg (about 5 mmol) was dissolved in 5.0 ml of 0.2 M N-ethylmorpholine acetate buffer, pH 8.0, and to this was added 1.6 g (about 11 mmol) of PGO-hydrate dissolved in 5.0 ml of the same buffer, plus 10 ml of methanol. The reaction mixture was kept at 25° for 20 h in the dark. The resulting precipitate was separated from the supernatant and washed with 30% methanol (3 ml x 2) and then water (3 ml x 3), and lyophilized. The yield was nearly quantitative. Calculated for C ^ H ^ N , (341.374): C 63.33, H 5.61, N 12.31. Found: C 63.78, H 5.20, N 12.34. Preparation and Characterization of GOarginine*—L-Arginine-HCl, 1 g (4.7 mmol) was dissolved in 0.27 M NaHCO3 and to this was added 5g (about 70 mmol) of GO-hydrate. The mixture was kept at room temperature for 1 h. To the resulting slightly yellowish reaction mixture was added 50 to 80 /JI each of methanol and ether. The oily material formed was washed several times with ether to get light brown crystals. This material was not purified further and the elementary analysis was not performed. A portion of the material was used for amino acid analysis before and after hydrolysis with 6 N HC1 at 110° for 24 h in an evacuated sealed tube. Another portion was incubated at different pH values (pH 0 to 12) at 25° for 20 h to investigate the stability of GO-arginine by amino acid analysis. Reaction of GO with L-Alanylglycine—A 0.1% 4

The derivatives formed by reaction of GO and MGO with arginine were tentatively named as GO-arginine and MGO-arginine, respectively.

Vol. 81, No. 2, 1977

aq. solution (20 (i\) of L-alanylglycine was mixed with 50 ft\ each of a 3 % GO aq. solution and 0.4 M NaHCO,, and the mixture was kept at 25°. At intervals aliquots were removed to determine the amount of the remaining peptide with an amino acid analyzer. L-Alanylglycine was eluted at or near the position of isoleucine. RESULTS Reactions of PGO with Amino Acids—Figure 1 shows the rates of reaction of various amino acids with PGO at different pH values. In general, the rate of reaction increased as the pH was increased from 5.5 to 8.0, and arginine was always lost most rapidly, followed by histidine and cystine. The other amino acids did not decrease significantly: their losses were in the range of 0 to 5 % in 24 h even at pH 8.0, with the exception of tryptophan, asparagine, glycine, glutamine, lysine, and methionine. The rates at pH 9.0 were quite similar to those at pH 8.0 except in the case of glycine. The amino acid analyses of these reaction mixtures indicated no new ninhydrin-positive amino acid derivatives. In the previous studies (i), di-PGO-L-arginine was prepared by direct reaction of arginine and PGO, followed by fractionation on a Sephadex G10 column. In the present studies, attempts were made to prepare the compound from N'-benzyloxycarbonyl-L-arginine as the starting material. The procedure was simpler than that used previously and the product was indistinguishable in properties from the previous preparation. The stoichiometry of the PGO reaction with guanidino group was also confirmed by preparation and analysis of a PGO-derivative of methylguanidine. Reactions of GO with Amino Acids—The rates of reaction of GO with various amino acids are shown in Fig. 2, and are fairly similar to those obtained with PGO. However, the rate of reaction of arginine was somewhat slower than with PGO, whereas that of lysine increased significantly. When histidine, 1-N-methylhistidine, and N°acetylhistidine were separately incubated with GO at pH 8.0 or pH 7.0 for 12 h under similar conditions, no significant change was indicated by the Pauly reaction (5). On the other hand, when examined by the manual ninhydrin method (6) and with the amino acid analyzer, histidine and 1-N-

398

K. TAKAHASHI