Fenton Chemistry

\’(>I.266, No. 26, Issue olSepternher 1.5. pp. 17201-17211, 1991 I’rinlrd in I f. S . A

Fenton Chemistry AMINO ACID OXIDATION* (Received for publication, March 28, 1991)

Earl R. StadtmanS and BarbaraS . Berlett From the Labf1rator.y o/ Biochemistry, National Heart, Lung, and Blood Institute, National Institutes of Health. Bethesda, Maryland 20892

Theoxidation of aminoacids by Fentonreagent involve the conversion of histidine residues toasparagine (H,O, Fe(I1)) leads mainly to the formation ofNH:, residues (5), of proline residues to glutamic semialdehyde and a-ketoacids, CO,, oximes, and aldehydes or carboxylic to pyroglutamic or glutamicacidresidues (6), of arginine acids containing one less carbon atom. Oxidation is residues to glutamicsemialdehyderesidues (6), and of all almost completely dependent on the presenceof bicar- theseandotheramino acid(especially lysine) residues to bonate ion and is stimulatedby iron chelators at levels carbonyl derivatives (6, 7 ) .Such oxidations are marking steps which are substoichiometric with respect to the iron in proteinturnover (1, 8-13); manyare implicatedin the concentration but is inhibited a t higher concentrations. accumulation of altered forms of enzymes during aging (14The stimulatory effect of chelators is not due merely 18),during oxidative stress (16), in neutrophil function (19), to solubilization of catalyticallyinactivepolymeric in various pathological conditions (ZO), and in ischemia-reforms of Fe(OH)3 nor to the conversion of Fe(I1) to complexes incapable of scavenging hydroxyl radicals. perfusion injury (21). Results of mechanistic studies are conH,Oe which are produced The resultssuggest that an iron chelate and another as sistent with the view that Fe(I1) and by the MFO systems undergo site-specific Fenton reactions yet unidentified form of iron a r e both required for maximal rates of amino acid oxidation. The metal ion- at metal-binding sites on the proteins to generate an active catalyzed oxidation of amino acids is likely a “caged” oxygen species (OH’, ferry1 ion, singletoxygen), which attacks side chains of amino acid residues at the metal-binding process,since the oxidation is not inhibitedbyhydroxyl radical scavengers, and the relative rates of site (4). In contrast to thesite-specific reactions catalyzed by metal oxidation of various amino acidsby the Fenton system as well as the distribution of products formed (espe- ions, the exposure of proteins to ionizing radiation leads to cially products of aromatic amino acids) are signifi- modification of virtually all amino acid residues, albeit cyscantly different from those reported for amino acid teine, histidine, tyrosine, methionine, and tryptophan resioxidation by ionizing radiation. Several iron-binding duesarepreferentialtargets (23-25). Curiously, the latter proteins, peptides, and hemoglobin degradation prod- three amino acidresidues are not common targets for the ucts can replace Fe(1I) or Fe(II1) in the bicarbonate- metal ion-catalyzed reaction, presumablybecause they are not dependent oxidation of amino acids. In view of their usually present at the protein metal-binding sites. Nevertheability to sequester metalions and their susceptibility less, results of studies on the oxidation of peptides (26, 23) to oxidation by H202in the presence of physiological and proteins (23-30) by ionizing radiation have contributed concentrations of bicarbonate, amino acids may serve significantly t,o our understanding ofhow these molecules an important role in antioxidant defense against tissue interact with oxygen radicals. Moreover, detailed studies of damage.

+

the oxidationof amino acidsby ionizing radiation have established that NH:, a-ketoacids, aldehydes, and hydrogen peroxide are among themajor products (24,31,32), andplausible mechanisms for the generation of these products have been Various enzymicand nonenzymic mixed-function oxidation proposed (23,24,31). Inview of the fact that the modification (MFO)‘ systems’ catalyze the oxidative inactivation of en- of proteins by ionizing radiation is at least qualitatively difzymes (1-4). It has been established that these oxidations ferent from those catalyzed by metal ions, it was of interest to investigate further the oxidation of amino acids by the * The costs of publication of this article were defrayed in part by Fenton system. the payment of page charges. This article must therefore be hereby The present studywas prompted by the report of Zs-Nagy marked “aduertisement” in accordance with 18 U.S.C. Section 1734 and Floyd (22) showing that the reactionof H,O, with Fe(I1) solely to indicate this fact. .$ To whom correspondence should be addressed Bldg.3, Rm. 222, complexes of ADP or ATP leads to the generation of OH’ National Institutes of Health, 9000 Rockville Pike, Bethesda, MD radicals which couldbe trapped by the spin trap 5,5-dimethyl20892. Tel.: 301-496-4096. 1-pyrroline-N-oxide (DMPO) and quantitated by means of ’ The abbreviationsusedare: MFO, mixed-functionoxidation; spectroscopy. Formation of the desferal, desferrioxamine; PRI, product recovery index; MCO, metal-electronspinresonance catalyzed oxidation; G(NH,,), number of NH., molecules produced per DMPO-OH’ spin adduct was quenched by the addition of 100 eV of radiation energy consumed; y(NH. Met MINUTES DESFERAL/Fe(ll) > His > Arg > Leu > Ser > Ala > Asp > Gly. In the metalrelative rates of FIG. 6. Dependence of leucine oxidation on the Fe(II)/des- catalyzed oxidation system used here, the feral ratio. Initially the main compartment of the double side arm oxidation of these same amino acids is quite different: Leu > Warburg vessels contained 23.5 mM NaHCO:+,50 mM leucine, and Ser > Ala > Gly > Arg > Trp > Met > Asp > His (Table111). various amounts of desferal as follows: 0, none; A, 40 pM; 0, 80 pM; (iii) The oxidationof amino acid by radiolysis is inhibited by 0 ,100 p ~A,; 110 p ~W,; 120 pM; X, 150 pM. After equilibration with OH’ radical scavengers (31), whereas these scavengers have 5% CO,, 95% N,,100 p~ FeS04 was added from one side arm, and then 30 mM H202was added from another side arm. Panel A shows little or no ability to inhibit the oxidation by MCO systems the time course of gases (CO, + 0 2 )which were formed during 30 (Table IV). This suggests that the oxidation of amino acids min. Panel E shows the concentrationof carbonyl compounds present by MCO systems is a caged process (see below). (iv) Theyield in thereaction mixtures after the 30 min of incubation at the indicated of ammonia in the decomposition of simple aliphatic amino chelator/Fe ratios. acids by radiolysis is inversely proportional to the numberof carbon atoms. TheG(NH,) values decrease from 2.3 to 0.3 as the number of carbon atoms is increased from 2 to 10 (31). This is attributed to the fact thatby increasing the lengthof the carbon chain, the numberof C-H bonds that aresusceptible to attack by OH’ is increased, and this leads to the generation of more kinds of products. In contrast, the yield of NH, in metal-catalyzed oxidation of aliphatic amino acids is relatively independent of the chain length (Table 111). It for the follows from Reactions 1-7 thataftercorrecting amount of H,O, consumed by disproportionation (i.e. 0, production) and by the secondary reactions (2 and 5 ) (CO, production), the amount of H,02 consumed in the oxidative deamination of amino acids is given by the expression (H20,)~= (HZO,), - (‘20,+ 20,)

00’~”

1

5

-L

10 15 CHELATORIFelllI

CHELATOR/Fe(ll)

FIG. 7. Dependence of leucine oxidation on the chelator/Fe ratio. A , reaction mixtures contained initially, 23.5 mM NaHCO,$, 100 H M FeSO.,, 20 mM H,O,, 50 mM leucine, and amounts of a,aor EGTA (0) dipyridyl (DZPYR,0),ADP (A), o-phenanthroline (0), t o yield chelator/Fe ratios as indicated on the abscissa. The FeS04 and H202were added aftergassing with the5% CO,, 95% N, mixture. After 30 min at 30 “C, the concentrations of carbonyl compounds weremeasured as described under “Experimental Procedures.” B, reaction mixture (0.4 ml) contained 23.5 mM NaHCOc3,100 p~ FeS04, 30 mM H,O,, 50 mM leucine, andconcentrations of desferal (0), EDTA (O),or diethylenetriaminopentaacetic acid (A)to yield chelator/Fe ratios as indicated on the abscissa. The data are the average of at least three separate experiments.To facilitate comparisons, the data were normalized with respect to the amount of carbonyl compounds observed at a cbelator/Fe ratio of 1.0. Otherwise conditions and procedures were as described in A .

where (H20Jt= total amount of H202consumed and (H202)* represents the amount H202consumed in the primary deamination reaction. The ammonia yield, y(NH3), is therefore given by the ratio (NH3)/(H20,) which should be equal to1.0 if the amino acid oxidation is due only to Reactions 1 and 4. By this criterion, it is evidentfrom the data in TableI11 that the yield of NH, in the oxidation of simple aliphatic amino acids is independent of thenumber of carbonatoms, i.e. y(NH3) = 1.0 for alanine, leucine, valine,and isoleucine. These results are therefore in contrast to those obtained by the radiolysis mechanism and are consistent with the view that the oxidation by MCO systems is a “cage”-type process in which amino acid-iron chelate complexes are reactive intermediates (see below). Themultiphasic response of amino acid oxidationto changes in the[chelator]/[ironsalt]ratio is not unique; similar effectshave been observedfor the peroxidation of lipids (44-48) and in thegeneration of OH’ by xanthine been suggested oxidase (49). Among other explanations, it has that stimulation of the oxidativeprocesses by EDTA and other chelators at [chelator]/[iron salt] ratios of less than 1.0 is due to an increase in the availability of soluble iron complexes (44, 47-49). This suggestion is based on the consideration that atphysiological pH values and the concentrations of iron saltscommonly employed (>75 p ~ in) in vitro studies, Fe(II1) would exist almost exclusively in the form of large insoluble Fe(OH)3polymeric complexesand would be unavailable for Fenton-type chemistry. Stimulation of the reaction by substoichiometric amounts of chelators was therefore at-

Fenton Chemistry: Oxidation AcidAmino

17208

nisms to explain the phenomenon are stilllacking. As noted by Gutteridge et al. (47), a complex formed by the binding of two metal ions as might occur a t low [chelator]/[iron] ratios could promote oxygen-catalyzed reactions, whereas 1:l complexes or differentkinds of complexes at high [chelator]/ [iron] ratios might be inactive or serve as inhibitors of oxidation reactions. In viewof thefactthatironcan form complexes with bicarbonate and amino acid as well as with EDTA and other chelators, the multiphasic response of amino acid oxidationtovariationsinthe[chelator]/[iron]ratio might reflect variations in the composition, concentration, and redox potentials (46, 53, 54) of various iron complexes that are formed under our experimental conditions. We reported earlier (55) that the Mn(I1)-catalyzed disproportionation of H 2 0 2is a third order function of the bicarbonate ion concentration andis reduced to a second order process by the presence of amino acids. It was proposed that in the absence of ADP 3 eq of HCO, react with Mn(I1) to form the catalytically active complex, but in the presenceof amino acid 1 eq of HCO, is replaced by the amino acid to yield an active complex of the composition, amino acid. Mn(11) . (HCO;),. The formation of a similar complex in which a chelator is also involved is suggested by the data in Fig. 1, showing that in theabsence of ADP thereis a sigmoidal responseof leucine oxidation to increasing concentrations ofHCO;, whereas in the presence of ADP a hyperbolic response is observed. This suggests that in the absence of ADP more than 1 eq of HCO; is involved in production of the catalytically active complex, whereasin the presence of ADP only 1eq is required. Although not shown, a double-reciprocal plot ( l / v versus 1/ Fe(I1) + OH’ + Fe(II1) + OH(10) HCO;) of the data obtained in the presence of ADP is linear, since the inhibitionof amino acid oxidation by high concen- from which an apparent dissociation constant of 13 mM can trations of iron (Fig. 9A, solid circles) is prevented if some be calculated for the binding of HCO; to form the putative ADP. Fe(I1). HCO:,. leucine EDTA iron complex is present in addition to the unchelatedcatalytic complex(possiblya complex). By monitoring spectral changes thatoccur in variron (Fig. 9A, solid triangles). ious reaction mixtures, we have obtained direct evidence for Considered together, the results summarized in Figs. 6-9 the formation of acomplexbetween chelator, Fe(II), and support the view that two forms of iron, the EDTA.iron complex and another form,possibly an amino acid chelate or amino acid (36).” Thus, when Fe(I1) is added to a mixture a bicarbonate complex, are required for optimal ratesof amino containing the iron chelator ferrozine and any amino acid, a acidoxidation. With substoichiometric levels of EDTA or bluecomplex (A,,, = 630 nm) of the composition, (ferrois desferal, both chelated and unchelated forms of iron would be zine):