Document wry9edbKY9epjY7wg1mO74NdB

Proteins THEJOURNAL 01 BIOLOGICAL CHEMISTRY Val. 268. No. 4. Issue ofF c b ~ a n6. pp. 2005-2008. 1991 Printed in U.S.A. ~JS law %. CATALASE PHYSIOLOGICAL CONSEQUENCES Earl R.StadtmanS: and Cynthia N. Oliver8 From the Laboratory of Biochemistry, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Marylnnd 20892 Detailed studies of the oxidative modification of glutamine >ynthetase (GS), in extracts of Escherichia coli (1, 3) and Klebsiella aerogenes (2, 4) established that the oxidation is dependent upon NAD(P)H, 02,and Fe(II1) or Cu(I1). The requirements for O2and an auxiliary electron donor (i.e. a donor other than GS itself) indicated that the oxidation is promoted by a mixed function oxidation (MFO)' type mechanism. In the meantime, it was found that many enzymes are readily inactivated by MFO systems (5-7) and that this renders them susceptible to proteolytic degradation by a variety .If exogenous and endogenous proteinases (1,8-12). It became e;.ident also that the oxidation of proteins is likely implicated in the killing of bacteria by neutrophils (131, and in the generation of altered (inactive or less active) forms of enzymes that accumulate during animal aging (6, 14-16) and under several pathological conditions (17-23).It is the intent of this review to summarize the results of studies on the mechanism of protein oxidation by MFO systems and the biological implications of such oxidations The Mechanism of Metal Ion-catalyzed Protein Oxidation The current view of how MCO' systems mediate protein oxidation is illustrated in Fig. 1. In this representation, the electron donor system is needed only to catalyze the reduction of O2to H202and of Fe(II1) to Fe(I1). Depending upon the electron donor system used, the reduction of O2can proceed by a two-electron mechanism yielding H20zdirectly, or by way of two sequential one-electron transfer processes leading first to superoxide anion followed by its dismutation to H202 and 0,. Similarly, the reduction of Fe(II1) to Fe(I1) can occur directly or via the intermediate formation of O,; which can -eact directly with Fe(II1) to form Fe(I1) and 0,. It is believed .hat the Fe(I1) then binds to a metal binding site on the protein (P) after which the protein. Fe(I1) complex reacts wjth the H20, to generate in situ an activated oxygen species (OH, ferry1 ion), which reacts with the side chains of amino acid residues a t the metal binding site. Among other modifications, some amino acid residues are converted to carbonyl derivatives (25, 26). After oxidative modification, the protein becomes highly sensitive to proteolytic degradation, and in $To whom correspondence should be addresad National Institutes of Health, goo0 Rockville Pike, Bldg. 3, Rm. 222. Bethesda. YD '20892. Tel.: 301- 496-4096. 0 Present address: Merck Sharp and Dohme Research Laboratories. Bldg. 16-100. West Point, PA 19486. 'The abbreviations used are: MFO. mixed function oxidation: MCO. metal- catalyzed oxidation: PBN, N-,ere-butyl-a-phenylnitrone: LDL. low density lipoprotein. 'To avoid confusion with mixed-function oxidmes (24). MFO systems are now referred to as metal-catalyzed oxidation MCO systems. FIG 1. Mechanism of enzyme oxidation and degradation. SOD. super. oxide dismutase: P. protem: Pa.,oxldized proteln. The symbols stunulation and inactivation. respectively. and 0 indicae the case of enzymes they are converted to catalytically inactive or less active, more thermolabile forms (14). Site-specificNature of the MCO-catalyzed Reactions According to Fig. 1, the MCO-catalyzed oxidation of proteins is a site-specific process involving the interaction of H,Oz and Fe(I1) at metal binding sites on the protein. The site-specific nature of the reaction is indicated by the following facts. ( a ) The inactivation of enzymes by MCO systems is relatively insensitive to inhibition by free radical scavengers (formate, ethanol, and mannitol) (5,6, 2 7 . ( b )Only one or at most only a few amino acid residues in a protein can be modified by MCO systems (25, 26) whereas almost all amino acid residues can be modified when proteins are subjected to free radicals obtained by radiolysis (28-30). (c) Most of the enzymes that are highly sensitive to modification by MCO systems require metal ions for catalytic activity. Therefore, they must contain a metal ion binding sire. ( d ) In the case of E. coli glutamine synthetase, the loss of catalytic activity correlates with the loss of a single histidyl and a single arginyl residue per subunit, both of which are situated in close proximity to one of two divalent metal binding sites on the enzyme (31.32); A plausible mechanism for the site-specific modification of a lysyl residue a t the metal binding site of a protein is illustrated in Fig. 2. In this mechanism the reduction of Fe(1II) (step a ) is followed by binding of the Fe(I1) to the enzyme (step b ) to form a coordination complex in which the c-amino group of a lysyl residue a t the metal binding site Serves as one of several ligands to which the Fe(I1) is bound. The H,02 produced by the reduction of 0, (step c ) may ream with Fe(1I) in the complex to form OH, OH-, and a Fe4III).enzyme complex (step d ) . The OH thus formed abtracrs a hydrogen atom from the carbon atom bearing the c-amino p u p to form a carbon-centered radical (step e), which then donates its unpaired electron to Fe(II1) in the complex to regenerate Fe(I1) and coincidently converts the c-amino group to an imino derivative (step f). Finally, the imino derivative undergoes spontaneous hydrolysis whereby S H i Fe(II)are released and an aldehyde derivative of the lysyl residue is generated (step g). The derivatized protein is thus rendered susceptible to proteolytic degradation. Upon acid hydrolysL 5-aminoadipic semialdehyde would be one of its pmducrr In this rep- resentation, the reaction of H202 with the Fe{II)-enzyme complex is shown to yield OH as a primary prcduct. Similar '1. Climent and R. L k i n e . personal c o m m u n m s n 2005 2006 Minireview: Metal-catalyzed Oxidation of Proteins \ 2 E+. ,<yiJO k - L a %.O wrn) H ~ O 06- U m t O ACmS FIG.2. Site-specific metal ion-catalyzed protein oxidation. TABLE I MCO s y s t e m that catalyze protein modificatwn 2. Xanthine oridase/..nthine/Fe(Iil)/O. 3. Cytochrome P-450/cytochrome P-450 reductase/Fe(llI)/O, 1. Cytochrome P-ISOAPutida redorin/P. redoxin reductase/ Fe(III)/NADH/O, 5. h r b a t c / O. J-F e ( I I I ) 6. Fe(II)/Oz 7. Fe(II)/H.O1 8. RSH/FeflIII/O. mechanisms in which (FeO)*+or (Fe(OH),)'+, or (FeOHI3++ OH- is the active intermediate might also be considered. Nevertheless, the reaction is perceived as a "caged" reaction in which the activated oxygen species does not escape but reacts preferentially with functional groups a t the metal binding site of the enzyme. This would account for the failure of radical scavengers to inhibit the reaction. It follows from the mechanisms depicted in Figs. 1 and 2 that any system capable of producing HzOzand of reducing Fe(II1) or Cu(I1) can provoke site-specific modifications of proteins possessing a metal binding site. Several enzymic and nonenzymic systems which have been shown to catalyze the oxidative modification of enzymes are listed in Table I. Perhaps the most physiologically relevant of these are those for which NADH or NADPH serves as the auxiliary electron donor, e.g. the NAD(P)H oxidases and the cytochrome P-450 systems. Of the nonenzymic MCO systems, those in which ascorbate and mercaptans serve as electron donors are particularly important. The ascorbate system, originally developed by Udenfriend et al. (33), has been considered a model for mixed function oxidases that catalyze the Oz-dependent hydroxylation of aromatic compounds (drugs) (24). In the meantime, it was found to promote the oxidation of lipids (34), nucleic acids (35), and proteins (25-27,36). Indeed, studies of Levine and co-workers on the oxidation of glutamine synthetase (25, 27) and amino acid homopolymers (26) by the ascorbate MCO system have contributed greatly to our understanding of the mechanism of site-specific metal ion-catalyzed oxidation of proteins. The biological significance of the RSH/Fe(III)/02 system seems assured by the discovery of a protein in yeast and in rat tissues that protects proteins from damage by this - MCO system (37, 38). Oxidation Marks Proteins forDegradation The proposition that metal-catalyzed oxidation marks proteins for degradation stemmed from the observations that the oxidized form of glutamine synthetase is degraded by cell-free extracts of E. coli more rapidly than is the native enzyme (1, 3). Confirmation of this hypothesis was obtained by Rivett (9, 10) and by Roseman and Levine (11) showing that highly purified preparations of neutral alkaline proteinases from rat liver and E. coli catalyze rapid degradation of the oxidized forms of glutamine synthetase but have little or no ability to degrade the unoxidized enzymes. The concept gained additional support from the studies of Davies and co-workers showing that the degradation of endogenous proteins in E. coli (39), red blood cells (40,41), and liver or heart mitochondria (42) is greatly enhanced following exposure of the cells to oxygen radicals or H202,and also by their studies showing that prior exposure of highly purified proteins to oxygen radical generation systems in vitro makes them highly susceptible to degradation by proteases in extracts of E. coli (39) or red blood cells (40). The discovery that carbonyl derivatives of some amino acid residues are among the products of oxygen radical damage t c proteins provides a means of assessing the extent of such damage under various physiological conditions. To this end, several methods have been developed for the quantitation of protein carbonyl groups (43). Protein Oxidation and Aging Catalytically inactive or less active, more thermolabile forms of enzymes accumulate in cells during aging (6,44).To test the possibility that the age-related changes are due in part to metal-catalyzed oxidation reactions, the concentr;. tions of protein carbonyl groups were measured in three different aging models, uu. in human erythrocytes, in cultured human dermal fibroblasts, and in rats of different ages. In the erythrocyte model advantage is taken of the fact that the density of erythrocytes increases with cell age. Therefore, cells of different ages can be separated from a single batch of blood by means of density gradient sedimentation. As shown in Fig. 3A, upper panel, the level of protein carbonyl groups in various fractions increased with cell density, Le. with cell age (14). Moreover, the levels of glyceraldehyde-3-P dehydro- FIG. 3. Age-dependent accumulation of oxidized protein and enzyme inactivation. A. the human eqthmcyts model: upper panel. the black and shuded bars refer to the protein carbonyl content of red cells from human donors 1 and 2. respectively:lower panol. the black, shadcd, and open born =fer to the relative catalytic activities of glyceraldehyde-3-P dehydrogenm (G-3PD). aspartats aminotransferase MAT),and phospbglycerate kinase (PGK), respectively. of human red cella B. the cultured fibroblast model shows the carbonyl content of protein from cultured human fibroblasts. C, the rat m d e k WDer wnel. the bars refer to the amount of carbonvl m u p a in protein f. Ai bvir extracts, mlddlepanol. the black and shoded&rs d e r td. the rela *e amount of matenal in rat liver extra& ( C R M )that cross-reacts anth anti&to purified glutamine synthetase (GS)and glueose-6-P dehydrogen- (G-& P D H ) respectively, lowerpanel. the black and shaded bars refer to the re]Suvr cntal$ic actwiaes of glutamine synthetnse and glucose-6-P dehydrogmfu* respectively The data in A and B are from experiments as described in O b e r et aL (14) The data in C are fmm Starke-Reed and Oliver (16).The Carb'JnY' content per mg of protein ( P ) IS expressed either an 'H per mg of Pmein obtained by reduction of carbonyl groups w t h ['HJBH, or DNPH pcr mg Of protein, where DNPH refen to the carbonyl content as a measured readion with DNPH. ge Pb co en ca vic ag wi Pr ex W1 ca Wl hi sa W1 in na of gr th D. G. w1 YO as Minireview: Metal-catalyzed Oxidation of Proteins 2007 genase (G-3-PD), aspartate aminotransferase (ATT), and of oxidized proteins which occurs during aging (Fig. 3, B and phosphoglycerate kinase (PGK) declined with cell density, in C) is by no means trivial. It represents a substantial increase confirmation of earlier studies showing the activities of these in the fraction of total protein that is oxidized. Based on the enzymes decrease with cell age. In the fibroblast model, the observation that on the average only one carbonyl group is carbonyl content of protein in cultured fibroblasts from indi- introduced per enzyme subunit and the assumption that the viduals of different ages and from individuals with premature average molecular weight of enzyme subunits is 50,000, it can aging diseases (patients with progeria or Werner's syndrome) be calculated that the amount of oxidized protein ranges from was measured using the 2,4-dinitrophenylhydrazoneassay 10% of the total protein in young animals to between 23 and procedure (14). As shown in Fig. 3B, there was an almost 30% in old animals (16).These are clearly minimal values exponential increase in the level of protein carbonyl groups since not all metal-catalyzed oxidation reactions yield car- with donor age over the range of 10-80 years. Moreover, the bonyl derivatives, uiz. histidyl, prolyl, and methionyl residues carbonyl content of protein in fibroblasts from individuals yield asparaginyl, pyroglutamyl, and methionyl sulfoxide res- with either progeria or Werner's syndrome was very much idues, respectively. higher than that in fibroblasts of normal individuals of the same age. In fact, the levels in fibroblasts from young donors Protein Oxidation and Disease with these aging diseases was about the same as that found ' n cultured fibroblasts from normal 80-year-old donors. Finally, as shown in Fig. 3C, upper panel, the carbonyl content of protein in hepatocytes from rats was found to increase gradually with animal age over the range of 3-20 months and then increased rapidly over the range of 20-26 months (16). Data in Fig. 3C, lower panel, show also that the activities of GS and glucose-6-P dehydrogenase (Glc-6-PD) decreased with animal age to values only about 50% of that found in young animals; moreover, the loss in enzyme activity was not associated with a comparable loss of antibody-specific immunoreactive protein (Fig. 3C, middle panel). Furthermore, he level of neutral alkaline protease(s) activity in hepatocytes .Is0 declines with animal age (Fig.4) suggesting that the age- dependent accumulation of altered enzyme forms is due partly to a loss of those proteases that degrade oxidized proteins. This possibility is supported further by an analysis of the data in Fig. 4 which shows a replot of some of the data in Fig. 3C. The shaded area in Fig. 4 represents the difference between catalytically active Glc-6-P dehydrogenase and total Glc-6-P dehydrogenase protein as measured by specific antibody titration. This shaded fraction therefore is a measure of inactive Glc-6-PD. As shown in the inset there is a linear inverse - dationship between the level of inactive Glc-6-PD and the -.vel of neutral alkaline protease activity. Taken together the data in Figs. 3C and 4 show that the level of oxidized protein increases with animal age and that the concomitant accumulation of altered Glc-6-PD is related to a loss of the proteinases that are responsible for the degradation of altered (oxidized?) protein. It should be emphasized that the 2-3-fold increase in levels Inflammatory Diseases-It is becoming apparent that MCO-mediated protein oxidation is an early indicator of tissue damage and that the formation of protein carbonyl derivatives is associated with pathological conditions both in humans and in animal model systems. Studies in this laboratory (45) have shown that activated neutrophils generate diffusible products (uiz. H20a O;,etc.) which in the presence of Fe(II1) will cause inactivation of enzymes within intact bacterial cells by an endocytotic independent mechanism. Moreover, during periods of oxidative burst some endogenous neutrophil proteins are converted to carbonyl derivatives and endogenous enzymes are inactivated. This may account for the fact that neutrophils lose activity and undergo lysis after activation. These results raise the possibility that recruitment and activation of neutrophils a t extravascular sites may provoke protein oxidation and thus contribute to the tissue damage that occurs during chronic inflammation. Indeed, Chapman et af. (21) have shown that the levels of protein carbonyl groups in the synovial fluid of patients with rheumatoid arthritis are higher than in patients with osteoarthritis. Atherosclerosis-It is now generally believed that metal ioncatalyzed oxidation of low density lipoprotein (LDL) by endothelial cells and the subsequent uptake of the oxidized LDL by monocytes/macrophages to form foam cells are important events in atherogenesis (for review see Steinberg et aL (46)). Most attention has been focused on the oxidation of the lipid moieties of LDL (47,48). However, the likelihood that protein oxidation is also involved is implied by the fact that modifications of apoprotein B include fragmentation and the loss of histidyl, lysyl, and prolyl residues; such changes are charac- teristic of MCO reactions (49). Ischemia Reperfcuion Tissue Damage-The proposal of McCord and co-workers (50) that oxygen free radicals are involved in ischemia reperfusion injury is supported by results of more recent studies with spin trapping techniques showing that free radicals are generated during the reperfusion phase (51, 52). A role of protein oxidation in ischemia reperfusion injury was verified by the finding that during reperfusion for 60 min following ischemia in the gerbil brain the level of protein 01 ob OX1012CO C-&POI ANIIGCU i o i 5 io AGE, MONTHS is ' FIG. 4. Relationship between protease activity and the accumulation ?I catalytically inactive glucose-6-P dehydrogenase protein. The data Are replots of data presented by Starke-Reed and Oliver (16). The eurues are as follows:1neutral alkaline protease activity; A, Glc-6-PD (G-6-PDH)activity: 0, protein that reacts with antibodies against highly purified Clc-6-PO. The shaded area is a measure of the catalytically inactive Glc-6-PD antigen which accumulated. The inset represents the relationship between the relative amounts of observed neutral protease activity and the amount of inactive Clc. 6-PD as determined by the difference between the total immunologically reactive Glc-6-PD protein and the catalytically active Glc-6-PD enzyme. carbonyl groups increased from 6 to 13 nmol/mg protein; coincidentally, the specific activity of glutamine synthetase declined to about one-half of its original value (52). That free radicals are involved in these changes is indicated by the fact that prior treatment of the animals with the spin trap N-tertbutyl-a-phenylnitrone (PBN) partially protected them against protein oxidation and loss of glutamine synthetase provoked by ischemia reperfusion. Neurologic Disorders-A role of protein oxidation in some neurological disorders is indicated by (a) the studies of KOMZ , 'I ,. . 2008 Minireview: Metal-catalyzed Oxidation of Proteins and Wiggins (S3) showing that myelin proteins are lost when rat myelin preparations are incubated with Cu(I1) and H202; (b)subsequent studies showing that when myelin preparations are incubated with XlCO systems, protein carbonyl groups are generated and the myelin is rapidly degraded by myelin proteases'; and (e) the studies of Mickel et al. (23) showing that rapid cormtion of vasopressin-induced hyponatremia in rats is associated with increased protein carbonyl content and myelinolysis. In addition, Murphy and Kehrer (19) have reported that the level of protein carbonyl groups is elevated in the pectoralis muscle of dystrophic chickens. Catarmtogenesis-Finally,Garland and co-workers (54,551 showed that there is a small but significant age-dependent increase in the protein carbonyl content of human lens, and that the level of oxidized proteins in cataractous lenses reaches values several times that of normal lenses. They showed also- that exposure of bovine crystallins to M C O systems in vitro produces physical changes similar to those observed in human crystallins in viuo. Conclmion It would appear from the above that the accumulation of oxidizedproteins may be an early indication of oxygen radicalmediated tissue damage, and in some pathological states may account for 50-70W of the total cellular protein (16).Since the intracellular level of oxidized proteins reflects the balance between the rates of oxidation and the rate of degradation of oxidized protein, the accumulation of oxidized protein is a complex function of the numerous factors that govern the synthesis and oxidation of proteins and the activities of various proteases that selectively degrade the oxidized forms. Except under special conditions (oiz. smoking or excessive exposure to high energy radiation), MCO reactions are likely responsible for most of the free radical damage. The processes that govern the levels of H20, and the availability of Fe(II1) and Cu(I1) may be fundamental to the etiology of various metabolic disorders. The level of H,Oz is dictated by the availability of 02 and the activities of numerous enzymic and nonenzymic MFO systems that generate H,O, on the one hand, and of those enzymes that catalyze the decomposition of H202 (peroxidases, catalase) on the other. Factors that control the availability of Fe(II1) and Cu(I1) are still poorly understood, but the release of Fe(I1) from ferritin and the release of Fe(II1) from Fe(II1).enzyme complexes by the degradation of hemoglobin as occurs under certain physiological and pathological conditions are the most likely sources. Factors that govern the levels and activities of multicatalytic proteases and other proteases that degrade oxidized protein are even less well understood. It is likely, however, that the loss of these protease activities is largely responsible for the accumulation of oxidized enzymes during aging and oxidative stress. The possibility that the loss of protease activity is also due to oxidative free radical damage deserves attention. REFERENCES 1. Oliver. C. N. (1981) The Regulation o/ Glutamine Synthetase Degmdafion Ph.D. thesis, The Johns Hopkins University 2. Levine, R.L.,Oljver. C. N.. Fulks. R. M.. and Stadtman, E. R. (1981) Pmc. Not1 Acad. SCLU.S.A. 78. 2120-2124 ~ ~~ 'J. Muller and C. N. Oliver, unpublished results. 3. Oliver. C. N.. Levine. R L.. and Stndtman. E. R (1981) in Merobolic Internnuenion a/ Enzymes (Holzer, H.. ed) pp. 259-268. Springer- Verlag. Berlin 4. Fulka. R M..and Sudtmnn, E. R (1985) B h c h i m Biaphys. Acta 843. 214-229 5. Stadtman. E. R. and Wittenberger, M. E. (1985) Arch. Biochem Bwphys. 239,379-387 6. Fucci. L. Oliver. C. N.. Coon. M. J.. and Stadtman. E. R. ( 1983) P m .NatL A d Sei U.S.A. 80.1521-1525 7. Chevion. M. (1988) Free Radical BioL 8; Med. 6.27-37 8. Farber. J. M.. and Levine. R L. (1982) Fed. P m .41.865 9. Rivett. A. J. (1985) J. B i d Chem 260, 12600-12606 10. Rivett, A. J. (1985) J. B i d Chem 260, 300-305 11. Roseman. J. E., and Levine. R L.(1987) J. BioL Chem 262.2101-2110 12. Daviea. K. J. A., and Goldberg,A. L.(1987) J . BioL Chem 262.8227-8234 13. Oliver, C. N. (1987) Arch Biochem. Blophya. 263,62-12 14. Oliver, C. N.. Ahn. B.-W.. Maeman. E. J.. Goldatein. S.. and Stadtman. E. m.R. (1987) J.BioL Chem 262,5488-5491 15. Garland, D. (1990) Eye Res. 60, in preas 16. Starke-Reed, P. E. and Oliver. C. N. (1989) Arch. Biochem. Biophys. 275. 559-567 17. Panhaaarathy, S., Wieland E.. and Steinberg, D. (1989) P m . NatL Acad. ScL U.S. A. 86.104-1050 18. Dicker, E.. and Cnllrbnum. A. I. (1988) FASEB J. 2,2901-2906 19. Mutphy M. E. and Kehrer J. P. (1989) Biochem J.260.359-364 20. Griffith;, H. R', Lunw. J., 6ee.C. k ,and Wilaan, R L. (1988) FEES Lett. 230.155-158 21. Chaoman. M. L. Rubin. 8. R.and Gracr. R W. (1989) J. RheumatoL 16, 15-18 22. Merritt. T. k. Boynen. B. R.and Northway, W. H..Jr. (1988) Broncho- p ~ l m D~yspphslo. Blackwell Scientificnear Book Medical Publish- ers. Inc.. Chicago 23. Mickel. H. S.. Starkc-Rccb P. E., and Oliver. C. N. (1990) Biochem. Biophys. Rea. Commun 172,92-97 24. Moson. H. S. (1957) Adu. E ~ y m o L19, 79-233 25. Levine. R L. (1983) J. B i d Chem. 268,11823-11627 26. Amici. A,. Levine. R L.Taia. L,and Stadtmnn. E. R (1989) J. BWL Chem 264.3341-3346 27. Levine R. L. (1983) J. B i d Chem 258, 11828-11833 28. Swalloh.. A. J. (1960) in Radiation Chemirt!y of Organic Compound (Swal- low, A. J.. ed)pp. 211-224. Pergamon Press, New York 29. G-n. W. M.. Jayko. M. E.. and Bennett, W . (1962) Radiution Res. 1 6 , 483-502 30. Davier. K. J. A. Debignore. M. E., and Lin. S. W. (1987) J. BWL Chem. 262,9902-9907 31. Farber. J. M.. and Levine. R L (1986) J. BioL Chem 261,4574-4578 32. Colombo C. and Villafmca, J. J. (1986) J.B i d Chem 261.10587-10591 33. Udenfri&d,'S., Clark, C. T., Axelrod. J., and Brodie, B. B. (1954) J. B i d Chem 2 0 8 , 7 3 1 34. Samokyszyn. V. M., and Aust, S. D. (1989) in Medical, Biochemical and Chemical Aspects a/ Free Radicab (Hayairhi, 0.. Niki. E., Kondo. M., and Yoshikawq T., eds)pp. 41-48, Elsevier. Tokyo 35. Chevion, M. (1988) Free Radicaf Biol. Q Med. 6.27-37 36. Deshpande. V. V..and Joshi. J. G. (1985) J. BIOLChem. 260,757-764 .--".37. Kim, K.. Rhee. S. G. and Stadtman. E. R (1985) J. E-1 Chem 260. 16!WA-1<997 -111. 38. Kim. I. H.. Kim. K.. and Rhee. S. G. (1989) Proc. Not[. A d . Sci. U.S . A 86, 601&6022 39. Davies. K. J. A,, and Lin. S. W. (1988) Free Radical BioL & Med. 5, 215- 223 40. Daviea, K. J. A. (1987) J. BWL Chem. 262.9895-9901 41. Pacifici. R E.. Salo. D. C., and Davies, K. J. A. (1989) Free Radicaf BWL & Med. 7,521-536 42. Marcillat, 0.. Zhang. Y.. Lm.S. W., and Davies, K. J. A (1988) Biochem J. 254.677-683 43. Levine. R L.. Garland D.. Oliver, C. N., Amici, A.. Climent, I., L e n s A.- G.. Ahn. B.-W., Shaltiel, S., and Stadtman, E. R. (1990) Methods En- zymol 1 8 8 , 4 6 4 4 7 8 44. Oliver. C. N., Fulks. R, Levine, R L.Fucci, L..Rivett. A. J.. Roseman, J. E.,and Stadtman. E. R (1984) in Molecular Basic o/Agin# ( b y , A. K.. and Chatteqee. B., eds)pp. 235-262, Academtc Pretia, New York 45. 46. Oliver. C. Steinberg. N. (1987) Arch Biochem D., ParthPsanrtby. S., C mBiwop, h'$.sE2.,5K3.h6o2o-. 72 J. C.. and Witzturr J. W. (1989) N. EngL J. Med. 320,915-924 47. Esterbauer, H., Joigens. G.,Quehenberger, 0.. and Koller. E. (1987) J . Lipid Res. 28,495-509 48. Steinbrecher U. P. (1987) J.BioL Chem 262,3603-3608 49. ~~ S ~ t e.i n b r e c h e r ; U. P.. Witaum. J. L.. Parthanamthy. S.,and Steinberg, D. (1987) A&r&c&su 1,135-1G. 50. Parks, D. A. Bulkley. G.B., Gunger, N.. Hamilton, S. R.. and McCord, J. M. (1982)'Gastroentomlogy 82,9-15 51. Floyd, R A. (1990) FASEB J . 4,2587-2597 52. Oliver, C. N.. Starke-Reed, P.. Stadtman, E. R, Liu, G.T., Carney. J. M., and Floyd R. A. (1990) P m .Not1 Acnd Sci U.S. A. 87,5144-5147 53. 54. Konat, G. W., and Wi Garland D.. Russell. !., gins, R. C. (1985) J. Neumhem. 46. 1113-1138 and Zigler, J. S., Jr. (1988) in Oxygen Radicals in Biology and Medicine (Simic. M. G.,Taylor. K. S., Wnrd, J. F., and van %.,Sonnta C.. eds)Vol. 49. pp. 347-353. Plenum P ~ MN.ew York 55. Garland. Zigler, J. S.. Jr., and Kinoshita. J. (1986) Arch. Biochem Biophys. 251, 771-776