Vol. 79
YOSHIHIKO MATSUO
2016 [CONTRIBUTION FROM T H E
DEPARTMENT O F PHYSIOLOGICAL CHEMISTRY, UNIVERSITY
O F CALIFORNIA SCHOOL O F M E D I C I N E ]
Pyridoxal Catalysis of Non-enzymatic Transamination in Ethanol Solution1 BY YOSHIHIKO ~IATSUO RECEIVED OCTOBER10, 1956
Transamination between pyridoxamine and a-keto acid, or amino acid and pyridoxal, readily occurs in absolute ethanol solution. Metal salts or a n elevated temperature are not required for the reaction. Under the same condition pyridoxal catalyzes transamination between amino acid and keto acid, thus mimicking the reactions catalyzed by transaminases. The reactions appear to proceed via the Schiff base intermediates.
Introduction Formation of Schiff base has been considered to be an intermediate step in reactions catalyzed by vitamin BGenzymes. Such a mechanism of reaction for transamination will involve a tautomerism of the aldimine (I) and ketimine (11) forms of the Schiff bases, as R-CH-COOH I
N
/I
R-C-COOH 11
S
CH
I
I1
Pyridoxal reacts in an alcohol medium with amino acids t o yield pyridoxylideneamino acids. Pyridoxylideneglutamic acid2 and pyridoxylidenealanine3 were isolated as yellow crystalline compounds from alcoholic solution. Brandenberger and Cohen2found t h a t a highly purified glutamateoxalacetate transaminase preparation from pig heart did not produce a-ketoglutaric acid when pyridoxylideneglutamic acid was used as the substrate, indicating t h a t the compound does not serve as an intermediate in the enzymatic transamination. This result would be expected, since pyridoxal phosphate, and not pyridoxal, is the coenzyme of transamination. Snell and co-workers have studied metal catalysis in a variety of non-enzymatic reactions involving amino acids and vitamin Be derivatives and emphasized the importance of metal chelation of the Schiff base intermediates. Eichhorn and Dawes5 confirmed the formation of such complexes from observation of the visible absorption spectra of the transamination reaction in aqueous solution. They reported t h a t the spectrum of the metal-pyridoxamine-pyruvate chelate became indistinguishable from that of the metal-pyridoxal-alanine complex upon standing. The present coiiimunication deals with a nonenzyniatic transamination system which functions a t room temperature in an alcohol medium. Addition of metal salts is not required for the reaction. 2p8
( 1 ) Airled b y research grants €tom t h e Kational Cancer I n s t i t u t e (Nos. 327 a n d 2327). t h e American Cancer Society (Met.-45A), a n d t h e Cancer Research F u n d s of t h e University of California. (2) H. Brandenberger a n d P. P. Cohen, Hela. Chinz. A d a , 36, 549
(1953). (3) D Heyl, S . A. Harris a n d K,Folkers, THISJOCRNAL, 70, 3429 (1948). (4) D . E. J I e t r l e r , &I. I k a w a a n d E. E. Snell, ibid., 7 6 , 648 (1954). ( 5 ) G . L. Eichhorn a n d J. W Dawes, ibid., 76, 5663 (1954).
Results (1) Formation of Schiff Bases in Ethano1.Absorption spectra of the yellow solutions prepared by mixing pyridoxal-HC1 and a-amino-nbutyric acid or pyridoxamine-2HC1 and a-ketobutyric acid in absolute ethanol are given in Figs. 1 and 2 . Spectra of pyridoxal-HC1 and pyridoxamine-2HC1 under the same condition are also given for comparison. Neither a-amino-n-butyrate nor a-keto-butyrate has significant absorption a t the concentration employed for the measurement M). of the spectrum (1.3 X It is clearly seen t h a t the spectra of the mixtures are entirely different from those of their components, indicating the formation of the Schiff bases. The difference between the spectrum of pyridoxal-a-amino-n-butyratecomplex and that of pyridoxaniine-a-ketobutyrate complex suggests the existence of the Schiff base tautomers I and 11. (2) Spectrophotometric Evidence for the Tautomerism of Schiff Bases.--A 0 01 116 aqueous solution of pyridoxamine and sodium a-ketobutyrate in 0.2 X phosphate buffer, p H 7.5, developed a faint yellow color only after standing 24 hours a t room temperature, while an equal concentration in ethanol solution gives an intense yellow color immediately. Rapid destruction of pyridoxylideneglutamic acid in aqueous solution has been reported previously.2 From this it was assumed t h a t addition of water to the alcoholic solution of Schiff base hydrolyzes the tautomers t o their most proximate components, i.e., (I) t o pyridoxal and amino acid, and (11) t o pyridoxamine and keto acid. This permits the course of the transformation in alcoholic solution of I1 into I to be followed by the 31)pearance of pyridoxal upon addition of water to the alcohol solution of the Schiff base. On this assumption the following cxperinient was undcrtaken. Aliquots were withdrawn a t iiitervals from the ethanol solution of the pyridoxaIiiirie-a-I\etobLityJI of the initial reactants), rate Schiff base diluted 60 times with aqueous 0.1 31 NaOII, and the absorption spectruni of the diluted solutio11 measured after 30 niiii. ( F i g . 3 ) . I t is seen that an absorption peak, which has its iiiaximutn a t 300 nip, increases in intensity with time.G It is well known t h a t pyridoxal, but not pyridoxamine, has an absorption a t this n.a\ e length in alkaline solution. Therefore, the results obtained indicate the appearance of pyridoxal in the aqueous solution of the pyridoxamine-a-ketobutyrate Schiff base, and (6) 0 . D . s ~measured at t h e 30th h r . of t h e experiment was sig-nilicantly l o a e r t h a n the value at t h e 16th hr. This is caused by t h ? reaction of pyridoxal a n d ethanol as discussed below.
April 20, 1957
NON-ENZYMATIC TRANSAMINATIOK
12017
Wave Length, mp
Wave Length, mp Fig. 1.-Spectrum of pyridoxdl-HC1 iii the presence and absence of a-amino-n-butyric acid: - - -, pyridoxalHCl (1.3 X 10-4 44); -, pyridoxal-HC1 and a-amino-nbutyric acid (1.3 X Arl each). Solvent, absolute ethanol in Fig. l a ; 3.3 X 10-3 Jrl XnOH in absolutc ethanol in Fig. lb.
-
Fig. 3.-Spcctrutii of an ethanol solution of pyridoxaniinc and a-ketobutyrate after dilution in aqueous 0.1 M NaOH. 10 pxnoles of pyridoxamine-2HC1 and sodium a-kctobutyrate were dissolved in 10.0 nil. of absolute ethanol, and the solution was shaken at room temperature. At intervals an aliquot was removed from the solution, diluted GO-fold with 0.1 ,II aqueous >aOH, and the absorption spectrum of the aqueous solution was determined after 30 min. of standing: 1, 1 hr.; 2, 5 hr. ; 3, 16 hr.; 4, 1.67 X l o w 4drl pyridoxamine2HC1 in 0.1 -If S a O H ; 5, 1.67 X lV4Jf pyridoxal-HC1 in 0.1 MNaOH.
thus provides evidence of the tautonierism of the Schiff base. No evidence for the formation of pyridoxal was obtained when pyridoxaniine and amino acid were incubated in an aqueous neutral buffer, or when pyridoxamine alone was incubated in ethanol. Spectral change with time observed in the ?TO330 mp region, where both pyridoxal and pyridoxamine have absorption, also indicated the decrease of pyridoxamine and increase of pyridoxal. X similar spectrophotometric study on the reverse of this experiment could not be carried out, because pyridoxal-HC1 reacts with ethanol, causing the decrease with time of the absorption a t 390 mp, measured in 0.1 171 NaOH, in the absence of aniino acid (Table I). The disappearance of pyridoxal in TABLE I REACTION O F PYRIDOXAL TYITII ETHASOL Optical density at 390 nip was determined on a Wfold 121 pyridoxal hydrodilution in 0.1 )If NaOH of 5.1 X chloride, which is prepared in absolute ethanol with the followiiig additions: ( l ) , no addition; (2j, 1 X Af XnOH; ( 3 ) , 1 X 31alaniriearid h-aOH. Incubatiou time, hr.
0 2
5 8.6
Wave L.ength, m p Fig. 2.-Spectruin of pyridosamiiie-2HCl in the presence of sodium a-ketobutyrate: - - - -, pyridoxamine2HCl (1.3 X X); -, pyridoxamine2HCl and a-ketobutyrate (1.3 X IOw4 ,IT each). Solveiit, absolute ethanol in Fig. 2a; 3.3 X -11 SaOIf in absolute ethanol it] Fig. 2h.
O.D.lS€ (1)
(3
0.I 5 0 ,144 1% ,104
0 1,711
,151
142 ,138
(3) 0 . 151
,150 150 ,146
the absence of amino acid is probably due to the formation of the hemiacetal and acetal of pyridoxal. The acetal formation is catalyzed by a trace of mineral acid and, hence, the reaction should be suppressed when the acid is neutralized. This effect may be seen from the table (column 2). =2ddition of an amino acid decreases the rate of disappearance
of pyridoxal (column 3), indicating that tlic formation of Schiff base competes for pyridoxal with the process of acetal formation. It also suggests t h a t the rate of disappearance of pyridoxal due to the tautomerism of Schiff base is slow, if it takes place a t all. (3) Formation of Amino Acids and Keto Acids by Non-enzymatic Transamination.-When pyridoxamine is incubated with an a-keto acid in absolute ethanol, an amino acid corresponding t o the keto acid employed is detected by paper chromatography. Alanine, norvaline, valine, norleucine, leucine, isoleucine, methionine, and glutamic, aspartic and a-amino-n-butyric acids were formed from the corresponding a-keto acids in this way* Similarly, a-ketobutyric acid is detected in the ethanol solution of pyridoxal and a-amino-nbutyric acid. The keto acid was identified by silica gel column chromatography and by paper chromatography of its 2,4-dinitrophenylhydrazone. Substances were detected in the ethanol solutions of other pyridoxylideneamino acids whose 2,4dinitrophenylhydrazones react positively t o the Friedemann-Haugen test.' a-ilmino-n-butyric acid was formed when aketobutyric acid and alanine were incubated in ethanol in the presence of pyridoxal. None of the reactants used here had a detectable contaniination of a-amino-n-butyric acid. An incubation of pyridoxal and a-ketobutyrate in ethanol produced no amino acid, proving that the pyridoxal was free from pyridoxamine. Thereforc, i t is evident t h a t under this condition alanine served as the amino group donor, the keto acid as the amino group acceptor, and pyridoxal as the amino group carrier, as represented in the equations
tioii of free reactants is alriiost negligible iii coniparison with t h a t of the complex. On this basis the results of the experinient summarized in Fig. 3 were interpreted t o indicate the transformation of the Schiff base from the ketimine type to the aldimine type, rather than t o signify a change in the concentrations of the free reactants. Estimation of the equilibrium point of the Schiff base tautomerism is complicated by the competing reaction of pyridoxal and ethanol, which will change the equilibrium by removing pyridoxal. If this difficulty could be circumvented, the system would offer a convenient means for estimating the approximate equilibrium constants to he expected in enzymatic transamination. Comparison of the spectrum in absolute iiietlianol of the potassium pyridoxylideneglutamate rcported by Brandenberger and Cohen2 with those of the similar derivatives of amino acids obtained in this work suggests t h a t thc spectrum characteristics reported by the above authors represented a mixture of spectra t o be observed under the basic and acidic conditions. Snell and co-workers4 suggested t h a t nietal chelation stabilizes the Schiff base in aqueous solution, and also enhances the tautomerism of the Schiff base by increasing the electron displacements at the a-carbon atom. Results of the present studies have demonstrated t h a t transamination proceeds readily at room temperature, and without metal catalysis, when an alcoholic medium is eniployed. The main effect of the alcoholic medium on the reaction inay be to provide a non-aqueous condition which favors the formation of the Schiff base intermediates. Once a substantial concentration of the Schiff base is formed in the reaction system, the tautomerism of I and I1 seems to occur without a chelation of metals. However, the alanine + pyridoxal = pyruvate pyridoxaminc ethanolic medium is so drastically different from pyridoxamine = a-ketobutyrate pyridoxal + a-amino-n-butyrate the aqueous physiological state, t h a t information obtained under such experiinental conditions docs Net: alanine a-ketobutyrate = pyruvate or-amino-n-butyrate not exclude the possible role of metal Chelation in This experinient was repeated with various com- the enzymatic system. But, on the other hand, it binations of amino and keto acids and the results is conceivable t h a t in the eniyiiiatic reaction the carboxyl group of the reacting amino or keto acids, were concordant. the phosphate group and other rcactivc groups of Discussion of Results the coenzyme will rcact, rrspectively, with the ac'l'lie stability constant, K , of pyridoxal phos- tive groups of the apocnzytne molecule. The nitrophate-amino acid Schiff bases, which is defined as gcii atom of the Schiff base may also react with apoenzyme. These bolldings will in effect stabi1ii.e [Schiff base] X [Water] the Schiff base intermediate, and iiiay be sufficient to [Amino acid] X [Pyridoxal phosphate] iiiaintain the plnnaiity of the coiijugated double has been estimated* to be of the order of l o 3to lo4. bond systcrii requisite for the electron displaccIt might be expected t h a t the pyridoxal-amino acid ments, without the help of rnctal chelation. Schiff bases also have stability constants of comSince an alcoholic medium greatly increases, withparable magnitude. If this is true, in a non- out nietal chelation, the concentration of Schiff aqueous medium, like the one used in this study, base over the value obtainable in aqueous solutions, the equilibrium greatly favors the formation of the effect of chelation on the tautorrierization of I Schiff base. Therefore, i t is likely that, in the and I1 may be studied directly in the non-aquc>ous alcoholic solution of the Schiff base, the concentra- medium, provided such chelation docs not cause (7) The 2,4-dinitrophenylhydrazoneprepared from the aqueous soluprecipitation of the reactantsQ
+ +
+
+
tion of pyridoxylidenehomoserine gave a positive Friedemann-Haugen test only when previously exposed to a solution of NaOH. This indicated that the product was the 2,4-dinitrophenylhydrazone of aketobutyrolactone (H. Hift and H. R . Mahler, J . Biol. Ckem.,198,901 (1952)). and thus demonstrated that transamination rather than deaininatioo of homoserine had occurred. (8) Y.Matsuo, THISJ O U R N A L , 79, 201 1 (1957).
(9) Results of a preliminary experiment on this problem indicated inhibited the transamination between that addition of C u + +or O f + + pvridoximine and n-keto acids in ethanol Whether this inhihitio~i W A cauqed ~ hy the direct action of the metal 1 0 1 1 on the tautomerism
