superoxide dismutase undergoes proteolysis and ...superoxide dismutase undergoes proteolysis and...

THE JOURNAL OF BIOLOGICAL. CHEMISTRY Vol. 265, No. 20, Issue of July 15, pp. 11919-11927,193O 0 1990 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.

Superoxide Dismutase Undergoes Proteolysis and Fragmentation following Oxidative Modification and Inactivation*

(Received for publication, March 5, 1990)

David C. Salo, Robert E. Pacifici, Sharon W. Lin, Cecilia Giulivi, and Kelvin J. A. Davies+ From the Institute for Toxicology and Department of Biochemistry, The University of Southern California, Health Sciences Campus, HSC-PSC 614-616, Los Angeles, California 90033

Red blood cells (RBC) are thought to be well pro- tected against oxidative stress by the antioxidant, cu- pro-zinc enzyme superoxide dismutase (CuZn SOD) which dismutates 0; to HzOz. CuZn SOD, however, is irreversibly inactivated by its product Hz02. Exposure of intact RBC to HzOz resulted in the inactivation (up to 50%) of endogenous SOD in a concentration-dependent manner. When RBC were exposed to 0; and HzOs, generated by xanthine + xanthine oxidase, an even greater loss of SOD activity (approximately 75%) was observed. Intracellular proteolysis was markedly increased by exposure to these same oxidants; up to a 12- fold increase with HzOz and a 50-fold increase with xanthine oxidase plus xanthine. When purified SOD was treated with Hz02, inactivation of the enzyme also occurred in a concentration-dependent manner. Ac- companying the loss of SOD activity, the binding of the copper ligand to the active site of the enzyme diminished with HzOz exposure, as evidenced by an increase in accessible copper. Significant direct fragmentation of SOD was evident only under conditions of prolonged exposure (20 h) to relatively high concentrations of H,Oz. Gel electrophoresis studies indicated that under most experimental conditions (i.e. l-h incubation) H202, O;, and HzOz + 0; treated SOD expe- rienced charge changes and partial denaturation, rather than fragmentation. The proteolytic susceptibility of HzOz-modified SOD, during subsequent incubation with (rabbit, bovine or human) red cell extracts also increased as a function of pretreatment with HzOz. Both enzyme inactivation and altered copper binding appeared to precede the increase in proteolytic susceptibility (whether measured as an effect of HzOz concentration or as a function of the duration of HzOz exposure). These results suggest that SOD inactivation and modification of copper binding are prerequisites for increased protein degradation. Proteolytic susceptibility was further enhanced by H202 exposure under alkaline conditions, suggesting that the hydroperoxide anion is the damaging species rather than HzOz itself. In RBC extracts, the proteolysis of HzOz-modified SOD was inhibited by sulfhydryl reagents, serine reagents, transition metal chelators, and ATP; suggesting the

* This work was supported by Grant ES 03598 from the National Institutes of Health/National Institute of Environmental Health Sciences (to K. J. A. D.). Some aspects of this work have been published in preliminary form (66). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accord- ance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ To whom all correspondence and reprint requests should be addressed: Institute for Toxicology and Department of Biochemistry, The University of Southern California, 1985 Zonal Avenue, HSC- PSC 614-616, Los Angeles, CA 90033. Tel.: 213-224-7542 or 213-224- 7895; Fax: 213-224-7473.

existence of an ATP-independent proteolytic pathway of sulfhydryl, serine, and metalloproteases, and peptidases. The proteolytic activity was conserved in a “Fraction II” of both human and rabbit RBC, and was purified from rabbit reticulocytes and erythrocytes to a 670-kDa proteinase complex, for which we have suggested the trivial name macroxyproteinase. In erythrocytes macroxyproteinase may prevent the accumulation of HzOz-modified SOD. In reticulocytes (and perhaps other cells which overexpress SOD as part of an oxidative stress response to 0; or HzOs) macroxyproteinase may form part of a closed-loop system for SOD turnover, regulated by the flux of 0; (or HzOz) to which SOD is exposed.

The superoxide anion radical (0;) is produced in most cells, including the red blood cell (1). Dismutation of 0, to Hz02 (and other reactive oxygen species) may subject the red blood cell (RBC)’ to significant oxidative stress. RBC are continuously exposed to 0; and H202 as a result of the autooxidation of hemoglobin (l-4), which can be further exacerbated by metabolism of various drugs and environmental agents (5-7). The antioxidant enzymes, superoxide dismutase (SOD) and glutathione peroxidase, as well as catalase at high H,Oz concentrations, form a substantial defensive network against oxidative stress. Several laboratories have reported that CuZn SOD is irreversibly inactivated by its product H202 (8-17), and it is well known that mature erythrocytes are unable to synthesize SOD (or other proteins) de nouo. Under normal circumstances a minor loss of SOD due to oxidative inactivation may not be of physiological significance. During exposure to oxidative stress, however, substantial SOD inactivation may occur, thus seriously compromising the antioxidant defenses of the RBC.

Oxidatively modified proteins have been shown to serve as preferred substrates for proteolytic enzymes in many cells (18-35). A novel intracellular proteolytic system in RBC appears to recognize, and preferentially degrade several oxidatively damaged proteins (18-28). We have proposed that selective proteolysis may prevent the intracellular accumulation of oxidatively modified proteins, which otherwise tend to form aggregates due to increased hydrophobic interactions and covalent cross-links (18-32). We have also proposed that

’ The abbreviations used are: RBC, red blood cells (used collectively for erythrocytes and reticulocytes); SOD, superoxide dismutase (the CuZn form of SOD was used in this work); M.O.P., macroxyproteinase; MgOAc, magnesium acetate; DTT, dithiothreitol; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; DIDS, 4,4’-diisothiocyanatostilbene-2,2’-disulfonic acid.

11919

by guest on March 22, 2020

http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

11920 Proteolysis and Fragmentation of Oxidatively Modified SOD

protein fragments’ are formed under certain oxidizing conditions and that the rapid proteolytic degradation of such fragments (which may have untoward biological effects) may represent a selective advantage (18-32). These previous pro- posals have largely been made on the basis of model experiments in which purified proteins were exposed to ‘OH (generated by y radiolysis of water). Based on such studies we hypothesized that CuZn SOD might be recognized and preferentially degraded in RBC, following inactivation by its product HzOz.

Preliminary work on this question suggested that H20,- modified SOD could, indeed, be degraded in erythrocyte extracts (26). Related work suggested that a 670-kDa proteinase complex (for which we have suggested the trivial name macroxyproteinase or M.O.P.) can preferentially degrade the ‘OH-modified forms of hemoglobin and albumin (27, 28). In the present report we have attempted to relate the Hz02- induced loss of SOD activity with structural changes in the enzyme active site (CL?+ binding) and increased proteolytic susceptibility. We have also attempted to characterize the proteolytic system which selectively degrades HzOz-modified SOD, including the proteinase responsible for initial recognition and cleavage.

EXPERIMENTAL PROCEDURES

Materials-All experiments described in this paper utilized a highly nurified bovine cunro-zinc-SOD (E.C. 1.15.1.1) obtained from DDI Pharmaceuticals Inc., Mountain View, CA (lot 186-72). The experiments were also repeated with a less pure bovine CuZn SOD from Sigma (S 8254), with essentially the same results (data not shown). SOD was radiolabeled by reductive methylation with [3H]formaldehyde and sodium borohydride as previously described (19). [3H] Formaldehyde (36 mCi/mol) was purchased from Du Pont-New Eng- land Nuclear. Stock solutions of H202 (Sigma) were made fresh daily and appropriate concentrations were determined spectrophotometri- tally (36), using an extinction coefficient of 39.4 M-‘.cm-i. Cyto- chrome c (Sigma) was succinoylated as described by O’Brien (37). Ammonium sulfate (Sigma) was ACS grade, all other chemicals were of reagent grade.

Preparation of Red Blood Cells, Cell-free Extracts, and “Fraction II”-Whole blood was obtained from human male adult volunteers, rabbits (Pel-Freez Biologicals, Rogers, AK; Product No. 31199), and cows. Erythrocytes were isolated by serial centrifugal washings as described previously (19-26). Some experiments were performed with intact cells in Krebs-Ringer phosphate buffer (23).

Cell-free extracts were prepared from RBC lysed by vigorous stirring (1 h) in 1.5 volumes of an hypotonic solution of 1 mM DL- dithiothreitol (DTT), as previously described (19-28). Unbroken cells, membranes, and organelles were removed by centrifugation. The resulting sunernatant, designated as “RBC lysate,” was extensively dialyzedagamst a membrane with a nominal molecular weight cutoff of 12.000. The dialvsis buffer contained 20 mM Tris-HCl CDH 7.0). 20 mM KCl, 1 mM magnesium acetate (MgOAc), 0.5 mM DTT, and 20% glycerol (v/v). Aliquots of the resulting material, designated as “cell- free extract,” were either used immediately for measurements of proteolysis or for protease isolation, or were temporarily frozen at -70 “C.

Fraction II was obtained by a procedure derived from that of Ciechanover et al. (38). Specifically, cell-free extracts were batch adsorbed to 0.5 volumes of pre-equilibrated DEAE-Sepharose CL-6B (Pharmacia LKB Biotechnology Inc.) for 1 h. Fraction I (mostly hemoglobin) was removed by washing the resin in a Buchner funnel with a buffer containing 20 mM Tris-HCl (pH 7.0), 20 mM KCI, 1 mM MgOAc, and 0.5 mM DTT, until the 415-nm absorbance declined to zero. The resin was then loaded onto a 5 X 5-cm column, and the

’ The term protein “fragmentation” refers to the direct breakdown of proteins by oxidative reactions. Such processes have been found to involve both main chain scission and side chain scission by nonen- zymatic mechanisms, which do not appear to involve peptide hydrolysis (18-22). In contrast, the terms “protein degradation” and “proteolysis” are used (interchangeably)-to refer to true peptide bond hydrolysis by proteolytic enzymes.

adsorbed material (Fraction II) was eluted (1 ml/min flow rate) with buffer containing 0.5 M NaCl, 20 mM Tris-HCI (pH 7.8), 20 mM KCl, 1 mM MgOAc, and 0.5 mM DTT.

Inactivation and Modification of Purified CuZn-SOD by H202- SOD (0.33 mg/ml, 10.3 NM) was inactivated by exposure to H202 in a buffer containing 50 mM Tris-HCl (pH 7.8), 5 mM MgOAc, and 0.5 mM DTT (24 “C). At appropriate times samples were removed and extensively dialyzed (using a membrane with a nominal molecular weight cut off of 6000-8000) against 2 X 800 volumes of distilled deionized water to remove excess Hz02 and potential SOD fragmentation products. In other experiments excess H202 was removed by the addition of catalase. Untreated SOD samples were similarly dialyzed, or incubated with catalase. Residual SOD activity (39), modification of active site copper binding, and proteolytic susceptibility were then measured.

Modification of the binding of the active site copper ligand was measured by a calorimetric technique utilizing bathocuproine disulfonate (17). To determine such modification, the following reagents were added sequentially at the final concentrations indicated: i.O% hvdroxvlamine, 0.02% bathocunroine disulfonate, and 1.0% ammonium acetate. Formation of bathocuproine-copper complexes was measured spectrophotometrically at 485 nm using an extinction coefficient of 1.0 x 10’ M-‘.cm-’ (17). Total copper was measured as described by Jewett et al. (17) following digestion of SOD with 7% HNO, and 1% H,Oz (at 105 “C). The 32-kDa native SOD dimer exhibited 1.9-2.0 mol of copper/m01 of enzyme (i.e. one copper atom per active site) following digestion; atomic absorption spectroscopy indicated 1.9-2.0 mol of zinc/m01 of enzyme.

Proteolytic Susceptibility of Hz02-modified SOD-The proteolytic susceptibility of control and HzOz-treated [3H]SOD was determined during incubation with red cell lysates, cell-free extracts, Fraction II, and further purified fractions, in a buffer containing 50 mM Tris-HCl (pH 7.8), 5 mM MgOAc, and 5 mM DTT. In lysates, and cell-free extract exneriments, 3.3 ucg of 13H1SOD was incubated with 3.9 ma of cell proteb in a final volume of 0:13 ml. Incubations were conducted in a rapidly shaking water bath at 37 “C, as previously described (18- 26). Similar incubations were performed with red cell Fraction II and further purified fractions (see Figure and Table legends for specific details). Following incubation, samples were placed on ice, precipitated with trichloroacetic acid (10% v/v final), and centrifuged at 3000 x g for 10 min. The trichloroacetic acid-soluble (super&ant) counts were determined by liquid scintillation, and percent [3H]SOD degradation was calculated as follows: % degradation = [(counts released - background counts)/(total counts L-background counts)] x 100. Where background counts were defined as acid-soluble radio- activity measured in the absence of RBC fractions (i.e. without proteolysis).

RESULTS

Superoxide Dismutase and Protein Degradation in Intact Erythrocytes-Incubation of isolated, intact RBC with increasing concentrations of HzO*, or with an O;/Hz02 generating system (xanthine + xanthine oxidase), for 30 min resulted in significant loss of CuZn SOD activity (Table I). Nearly 25% of the SOD activity was lost when RBC were exposed to 3.0 mM HzOz, and more than 50% inactivation was observed following exposure to 15 mM H202. While the loss of activity was clearly dependent upon Hz02 concentration, a much greater loss of SOD activity (70%) resulted when RBC were exposed to xanthine + xanthine oxidase, which generated both 0; and H202. The maximum concentration of H202 generated in our xanthine + xanthine oxidase experiments (assuming complete dismutation of 0;) would be approximately 1 mM. These results may indicate that dismutation of 0; at the SOD active site over a prolonged period (even at low concentrations) may be more effective than Hz02 in inactivating SOD. Alternatively, our results may reflect differential effects of continuous flux versus bolus additions of Hz02 or 0;.

Exposure to either Hz02 or xanthine + xanthine oxidase also markedly increased intracellular proteolysis, as evidenced by the formation of free alanine (Table I). Alanine is not

synthesized de novo or by metabolic interconversion of amino


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Proteolysis and Fragmentation of Oxidatively Modified SOD

TABLE I TABLE II

Proteolysis of intact erythrocytes following flux or bolus exposure to Inactivation of superoxide dismutase and increased proteolysis following exposure of intact erythrocytes to H202 or nanthine oxidase oxidants

Washed human erythrocytes were preincubated as a 10% (v/v) suspension in order to deplete free amino acids as previously described (23-25). The preincubated cells were then suspended to a final concentration bf 5% (v/v) in pH 7.4 Krebs-Ringer phosphate buffer (143 mM NaCl, 5.7 mM KCl. 1.4 mM M&L. 18 mM sodium phosphate, 6 mM glucosej. Aliquots (i ml) were&&bated in a shaking-wate; bath at 37 “C for 30 min in the presence or absence of H,O, or 1 mM xanthine + 0.2 units/ml xanthine oxidase. For measurements of SOD activity, samples were placed on ice following incubation and the cells were immediately lysed by addition of 1.5 volumes of distilled water (4 “C) containing 0.5 mM dithiothreitol. Membrane fractions were removed by centrifugation for 20 min at 18,500 x g. SOD activity in the supernatants was determined by the ability to inhibit the reduction of succinoylated cytochrome c by 0; (generated by xanthine and xanthine oxidase (39)). One unit of SOD activity is defined as the amount of SOD which inhibits the rate of reduction of cytochrome c by 50% (39). Proteolysis was assessed by formation of free alanine from previously intact cellular protein. Alanine was measured fluo- rometrically (in neutralized perchlorate-treated cell supernatants), by the (alanine-dependent) reduction of NAD+ to NADH catalyzed by (added) alanine dehydrogenase, as previously described (23). Val- ues are means + S.E. of three independent determinations. Similar results were also obtained with rabbit erythrocytes (data not shown).

Washed human erythrocytes were preincubated as 10% (v/v) sus- pensions in order to deplete free amino acids as previously described (23-25). The preincubated cells were then suspended to a final concentration of 5% (v/v) in pH 7.4 Krebs-Ringer phosphate buffer (143 mM NaCl, 5.7 mM KCl, 1.4 mM MgC&, 18 mM sodium phosphate, 5 mM glucose). Aliquots (3 ml) were incubated in a shaking water bath at 37 “C for 30 min in the presence or absence of various additions. Where indicated, xanthine (2 mM) and xanthine oxidase (nominally 0.001 units/ml) were added to produce 0; and H,O, at a rate of 1 nmol/min each, as experimentally determined by the method of McCord and Fridovich (39) and Boveris et al. (40), respectively. Total production amounted to approximately 90 nmol of 0; and 90 nmol of H202 during the 30-min incubation. Other cells were incu- hated with glucose (10 mM) and glucose oxidase (nominally 0.0155 units/ml) to generate HzOz at a rate of 1 nmol/min (experimentally determined), providing a cumulative exposure of 90 nmol of H,O, during the 30-min incubation. To best approximate the xanthine + xanthine oxidase and the glucose + glucose oxidase experiments with a bolus addition of Hz02 we. therefore, added 90 nmol of H202 in the 3-ml incubation volume (i.e. 30 pM final concentration). The anion channel blocker, 4,4’-diisothiocyanatostilbene-2,2’-disulfonic acid (DIDS), was used in further experiments with xanthine + xanthine oxidase to inhibit 0; entry into the red cells. For these experiments, RBC samples were preincubated in the presence of DIDS (1.67 pg/ ml) for 15 min, washed, and then resuspended in Krebs-Ringer phosphate buffer for incubation with xanthine + xanthine oxidase as above. Proteolysis was assessed by formation of free alanine from previously intact cellular protein. Alanine was measured fluoromet- iically (in perchlorate-treated cell supernatants), by the (alanine- deDendent) reduction of NAD+ to NADH catalvzed bv (added) alanine dehydrogenase, as previously described (23). values ark mea& f SE. of three independent determinations.

Addition SOD activity Proteolysis

units/mg red cell Hb nmol &nine/ml RBC

Control 108 + 7 0.1 + 0.01 3.0 mM HZ02 80 + 3 0.5 + 0.01 15.0 mM HZ02 56 + 9 1.2 + 0.02 Xanthine oxidase plus 28 f 4 5.1 + 0.06

xanthine

acids in erythrocytes, therefore the net production of this amino acid can only occur via protein degradation (23). As with SOD inactivation (see above) xanthine + xanthine oxidase was more effective (on a molar basis) in inducing proteolysis than was H,O,. Importantly, neither xanthine nor xanthine oxidase alone had any effect on SOD inactivation or proteolysis, and heat-denatured xanthine oxidase (+ xanthine) was also without effect (data not shown).

The observation (Table I) that an 0, + H202 flux (generated by xanthine + xanthine oxidase) induced greater proteolysis than did a “bolus” addition of H,O, led to additional experiments. When RBC were exposed to a 30 pM bolus addition of Hz02, a 2-fold increase in protein degradation resulted (Table II) as compared to untreated red cells. In contrast, when red cells were exposed to an equivalent H202 flux generated by glucose + glucose oxidase, proteolysis was increased some 15fold. Thus it would appear that a flux of H202 is a more efficient way to induce protein damage and degradation than is a bolus addition of H202.

The largest increase in proteolysis (21-fold) shown in Table II was observed when RBC were exposed to xanthine + xanthine oxidase (to generate both 0; and H202). The xanthine + xanthine oxidase system was adjusted to generate an H202 flux equal to that produced by glucose + glucose oxidase (1 nmol/min). The greater effectiveness of the xanthine oxidase system might, therefore, be due to the additional generation of 0;. To test this hypothesis we repeated the xanthine oxidase exposure with RBC which had been preincubated with the anion channel blocking agent 4,4’-diisothiocyanatostilbene-2,2’-disulfonic acid (DIDS), which prevents 0; entry across the RBC membrane. The addition of DID.5 decreased the proteolysis induced by xanthine oxidase to a level comparable to that seen with glucose oxidase (Table II), indicating that 0; does, indeed, contribute to the proteolytic response. Although we are unable to determine whether 0; acted directly or indirectly (i.e. by reacting with H202 to generate

Addition

None 30 PM H,O, bolus Glucose oxidase + glucose Xanthine oxidase + xanthine Xanthine oxidase + xanthine

+ DIDS

Proteolysis IIlCE%%3

nmol alanine/ml RBC -fold 0.1 + 0.01 0.2 + 0.01 2 1.5 * 0.10 15 2.1 * 0.20 21 1.6 + 0.10 16

‘OH) in inducing proteolysis during exposure to xanthine oxidase, these results indicated that 0; (the substrate of SOD) was worthy of greater study.

Physical modification of SOD following exposure to either Hz02 or 0; was examined by polyacrylamide gel electrophoresis (PAGE) under both nondenaturing and denaturing (sodium dodecyl sulphate (SDS)) conditions, as well as by isoelectric focusing (Fig. 1). Nondenaturing gels (Fig. lA) clearly demonstrated the formation of several bands following oxidative modification of SOD. Interestingly, exposure to either H202 or 0; resulted in the formation of at least 6 discrete protein bands demonstrating that 0; is, indeed, able to modify the enzyme in a manner similar to H202. The results of Fig. lA could be explained by charge changes, or by direct fragmentation of the protein. The possibility of direct fragmentation of the protein was, however, excluded by running SDS- PAGE gels (Fig. 1B) where lower molecular weight protein bands were not observed. Furthermore, isoelectric focusing gels (Fig. 1C) indicated that charge changes were at least partly responsible for the distinct bands observed in native gels (Fig. lA). These data would strongly suggest that SOD undergoes early partial denaturation upon exposure to oxidative conditions, thereby increasing the potential for rapid degradation of the modified protein. Additionally, the discrete bands seen with oxidative exposure may suggest that the reaction of H,O, or 0; with SOD may not be random, but rather may involve site-specific modifications, perhaps dic-


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/


FIG. 1. Polyacrylamide gel electrophoresis of SOD following exposure to oxidants. Samples of SOD (0.33 mg/ml) were exposed to various concentrations of H,02, as indicated (1 h at 24 “C), as described under “Experimental Procedures.” Additional SOD samples were exposed to a flux of 0; at a total dose of approximately 58 nmol of O.;/nmol of SOD by 100 k&d ““Co radiation (Gammacel600, Atomic Energ. of Canada, Inc.) in 10 rnM formate (pH 7.8) saturated with 100% O,, as previously described (19). These exposure conditions resulted in a cumulative exposure to 0; of approximately 594 nmol in a volume of 1 ml. Were all 594 nmol to dismutate, the maximal H,O, exposure would only have been 0.3 mM H,O,; an exposure which produced no detectable effects (data not shown). Panel A shows the results of (20% T? 2% C) nondenaturing PAGE. Panel B shows the results of samples run on (20% T, 2% C) denaturing SDS-PAGE following boiling for 5 min at pH 8.0 in SDS treatment buffer (10 mM Tris-HCI, 1 mM EDTA, 2.6% SDS, .i.O% /s-mercaptoethanol). I’nncl (‘ shows the results of samples run on an isoelectric focusing pel with a pH range of 4-6.5. All gels were run on an automated electrophoresis system (PhastSystem, Pharmacia) and stained with silver stain.

tated by the initial reaction at the active copper-binding site. Inactivation, Modification of Copper Binding, and Proteo-

lytic Susceptibility of Purified CuZn-SOD Following Exposure to H&&-The inactivation of purified CuZn SOD by H,O, is irreversible and appears to depend on the concentration of HrOz, as well as the pH and the duration of exposure. A progressive loss of SOD activity with increasing H,Or concentrations is shown in Fig. 2A. Some 25% of the SOD activity was lost following exposure to 1.5 mM H,O, for 1 h, while more than 90% of the SOD was inactivated following a l-h exposure to 30 mM H,O,. The binding of the copper ligand to the active site of the enzyme was also affected by HLO, exposure in a concentration-dependent manner as evidenced by a gradual increase in measurable copper (Fig. 2.4). Follow- ing exposure to 1.5 mM HzOz, 16% of the active site copper

0 IQ 20 30 H,Qz h,H)

FIG. 2. H,O, concentration effects on the inactivation, active site copper binding, and proteolytic susceptibility of SOD. [‘LH]SOD was incubated at 24 “C for 1 h (pH 7.8) with the concentrations of H,O, indicated. Following incubation. samples (both H,O, treated and untreated) utilized to determine inactivation and proteolytic susceptibility were extensively dialyzed against a membrane with a nominal molecular weight cut off of 6000-8000. Catalase c.5 p,-) was added to samples used for the determination of copper reactivity in order to decompose any possible remaining H,O,. Percent inactivation (0) and percent of total copper (0) which became accessible to the bathocuproine disulfonate reagent are shown in panel .A. Percent SOD degradation during subsequent incubation with bovine erythrocyte ext.racts (/I) and rabbit erythrocyte extracts (A) is shown in panel H. Measurements were performed as described under “Ex- perimental Procedures.” Values shown are means f S.E. of three independent determinations.

became accessible to assay by the bathocuproine disulfonate reagent. Nearly 60% of the total copper reacted with the bathocuproine reagent following exposure to 30 mM H,O,. These results suggest that the catalytically essential copper ligand (which is normally concealed within the tertiary struc- ture of the enzyme) becomes exposed following HrO, incubation due to unfolding or denaturation of the protein. Several studies have demonstrated that a single histidine residue, located at the active site, is modified by HrOl exposure (8-10, 12,13). Oxidative modification of this amino acid residue may trigger the denaturation of SOD and the exposure of the copper ligand to the bathocuproine reagent used in assaying copper.

The proteolytic susceptibility of SOD during subsequent incubation with red cell extracts also increased as a function of pretreatment with H,O,. Untreated (native) SOD was highly resistant to proteolysis, in comparison with other proteins (e.g. hemoglobin and albumin) which we have studied in red cell extracts (18-25). Degradation of SOD during incubation with cell-free extracts from bovine erythrocytes increased by 32-fold following a l-h pretreatment of the enzyme with 1.5 mM H.‘O.‘, and increased almost 70-fold following pretreatment with 30 mM H,O, (Fig. 2B). Similarly, rabbit erythrocyte extracts exhibited a 33-fold higher degradation of SOD pretreated with 1.5 mM H,O, and an almost loo-fold higher degradation of SOD pretreated with 30 mM H,O, (Fig. 2B). It is important to note that the absolute values for percent degradation of [“HISOD in Fig. 2B (and all subsequent Figures and Tables) were experimentally manipulated. We chose [“HISOD and cell extract concentrations which caused less than 10% protein degradation in order to avoid problems associated with substrate limitation. Thus, absolute values for SOD inactivation or altered copper binding should only be compared with relative increases in proteolytic susceptibility.

The duration of SOD exposure to a given concentration of HzO:, also affected enzyme activity, active site copper binding, and proteolytic susceptibility. After a 2-h incubation in 1.5 mM HzOz, SOD was more than 50% inactivated, and more than 95% inactivation was observed following a 20-h exposure (Fig. 3A). More than 30% of the total copper became acces-


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Proteolysis and Fragmentation of Oxidatively Modified SOD 11923

alkaline pH. We now demonstrate that the proteolytic susceptibility of SOD (during subsequent incubation with human or rabbit erythrocyte extracts) also increases as a function of the pH of pre-exposure to H,Oz (Fig. 5). The results of Fig. 5 provide further evidence for a strong relationship between oxidative inactivation/modification and proteolytic susceptibility.

FIG. 3.

T,m Chrd

Effect of duration 01 13H1SOD was incubated with 1.5 mM H202 at 24 “C for up to 20 h (pH 7.8). At the times indicated, aliquots were removed and treated as described in the legend to Fig. 1. Percent inactivation (0) and percent of accessible copper (0) are shown inpaneld, and degradation (A) during subsequent incubation with bovine erythrocyte extract is shown in panel B. Values shown represent means f SE. of three independent determinations.

Characterization of SOD Degradation in RBC Extracts- Although the RBC extract degradation of SOD increased with

TlrnP h-6) the pH of H202 exposure (Fig. 5), the pH optimum for deg- 30D exposure to 1.5 mM HzOz. radation (following preincubation at any pH) was found to be

7.8 (data not shown). This result is, perhaps, not surprising since we previously demonstrated that the pH optimum for the degradation of oxidatively denatured albumin (in RBC extract) is also pH 7.8 (19-22). Transition metal chelators (EDTA, %hydroxyquinoline), serine reagents (phenylmethylsulfonyl fluoride, diisopropyl fluorophosphate), and sulfhy- dry1 reagents (N-ethylmaleimide, PCMPS) were all effective

FIG. 4. Normalized results for SOD inactivation, active site copper binding, and proteolytic susceptibility following exposure to HzOz. Since absolute values for proteolysis in Figs. 1 and 2 are experimentally manipulated (extract concentration, SOD concentration, etc.) to yield values of less than lo%, enzyme inactivation, copper binding, and proteolysis can only be directly compared using normalized results. In this figure the data of Figs. 1 and 2 are represented as percentages of the maximal responses observed. Sym- bols used (in both panels A and B) are as follows: 0, SOD inactivation; A, accessible copper; q , degradation during subsequent incubation with bovine erythrocyte extracts. Values shown are means + S.E. of three independent determinations.

inhibitors of the degradation of H202-modified SOD in RBC extracts (Table III). These results suggest the possible involvement of metallo-, serine-, and sulfhydryl proteases/peptidases in the proteolytic pathway. Similar results were previously obtained for the degradation of other oxidatively modified proteins in RBC extracts (19-28). The inhibition of proteolysis by hemin is unexplained, but the ineffectiveness of leupeptin indicates that calcium-activated thiol proteases and lysosomal proteases are not involved (24).

Although several modified proteins can be degraded by an ATP/ubiquitin-stimulated proteolytic pathway in reticulocytes (41-43), this pathway appears to decline as reticulocytes mature into erythrocytes. In our hands the degradation of HBOz-modified SOD was actually slightly inhibited by the addition of ATP (+MgCl,) to rabbit or bovine erythrocyte extracts (Fig. 6). It is possible that ATP may have promoted the activity of a residual ATP-stimulated proteolytic pathway (against other protein substrates) which shares a common pool of peptidases. Peptidase activity may be rate-limiting when more than one proteolytic pathway is in operation. Whatever the actual explanation for our results, it seems clear that HzOP-modified SOD must be degraded by an ATP- and ubiquitin-independent pathway in RBC.

sible to the bathocuproine reagent after exposure to 1.5 mM

H202 for 2 h, and approximately 45% of the total copper was assayable when exposure was extended to 20 h (Fig. 3A). Proteolytic susceptibility during subsequent incubation of SOD with bovine erythrocyte extracts also appeared to depend upon the length of pre-exposure to a given concentration of H,Oz. After a preincubation period of 2 h with 1.5 mM H202, a 36-fold increase in proteolysis occurred, with a further increase to loo-fold following 20 h of preincubation (Fig. 3B).

Purification of the SOD-degrading Activity of RBC-As shown in Fig. 7, a Fraction II (28, 38) prepared by DEAE chromatography of human erythrocyte extracts exhibited

. 20

6 s -1.5. 0 b

$ 1.0.

j-J 0 i

N

as- i

When the data of Figs. 2 and 3 are re-evaluated or normalized as percentages of maximal responses (necessary because of the vastly different absolute scales of each response) it is evident that a rough correlation exists between the percent loss of SOD activity, the modification of active site copper binding, and the proteolytic susceptibility of HPOP-modified SOD (Fig. 4). Although such results do not demonstrate a causal relationship, these data are consistent with the proposal that SOD inactivation and modification of copper binding are prerequisites for increased proteolytic susceptibility. The experiments reported in Figs. 2-4 were repeated with heat-inactivated red cell extracts and no degradation was observed. Thus, the degradation reported in Figs. 2-4 repre- sents a true enzymatic activity. The inactivation of SOD by HzOz has also been observed to increase with increasing pH during H202 exposure (8). Blech and Borders (8) and others (15) have proposed that H202 must dissociate into its conjugate base, the hydroperoxide anion (HO;), in order to inacti- vate SOD, and the dissociation of Hz02 is clearly favored at

ao’ 6 8 10 ptl of H,O, Exposure

FIG. 5. Effect of HzOz exposure pH on the proteolytic susceptibility of SOD. [3H]SOD was incubated with 1.5 mM H,Oz at the pH values indicated for 30 min, and extensively dialyzed against 50 mM Tris-HCl (DH 7.8). 5 mM MeOAc. and 0.5 mM DTT. The treated protein was then incubated in the presence of rabbit (0) or human (0) erythrocyte extracts (60 min at 37 ‘C, pH 7.8) for proteolysis measurements, as described under “Experimental Procedures.” The H20, exposure buffers used were malatk (pH 4-5), Tris malate (pH 6-7), Tris-HCl (pH 7.5-g), and sodium borate (pH 10). Values shown are means + S.E. of three independent determinations.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Proteolysi-s and Fragmentation of Oxidatively Modified SOD

TABLE III Effects of potential protease inhibitors on the degradation of Hz02

modified SOD in erythrocyte extracts [3H]SOD was exposed to 1.5 mM H,O, for 60 min (24 “C, pH 7.8)

and extensively dialyzed prior to incubation with cell-free extracts of bovine erythrocytes (see “Experimental Procedures”). Erythrocyte extracts were first preincubated for 10 min with potential protease inhibitors. Appropriate control experiments were performed with inhibitor solvents (ethanol and dimethyl sulfoxide) whose effects have been subtracted where necessary. H20Q-modified [3H]SOD was then added to the extracts for 60-min incubations at 37 “C. Degradation was measured as described under “Experimental Procedures.” Values represent percent inhibition of the degradation of HoOz-modified SOD and are the means of three independent determinations for which S.E.s were less than 10%. Similar results were also obtained with cell-free extracts from rabbit erythrocytes and reticulocytes, and human erythrocytes (data not shown).

Addition Inhibition of degradation

EDTA (10 mM) SHydroxyquinoline (1 mM) Hemin (0.2 mM) Phenylmethylsulfonyl fluoride (5 mM) Diisopropyl fluorophosphate (5 mM) N-Ethylmaleimide (1 mM) p-Chloromercuriphenylsulfonic acid (2 mM) Leupeotin (0.1 mM)

% 42 53 76 81 85 76 95

6

IO 20 30 IO 20 30 H,O, (MM) H,O, hm

FIG. 6. Effect of ATP on the degradation of HzOz-treated SOD. [3H]SOD was exposed to various concentrations of H202, as indicated, for 1 h at 24 “C (pH 7.8) and extensively dialyzed (see “Experimental Procedures”). Treated (and untreated) samples were then incubated (60 min at 37 “C, pH 7.8) with either rabbit (panel A) or bovine (panel B) erythrocyte extracts, in the presence (0) or absence (0) of ATP + MgCl, (5 mM each). Proteolysis was measured as described under “Experimental Procedures,” and all values shown are means + S.E. of three independent determinations.

comparable selectivity in the degradation of HzOs-modified SOD to that seen with human, rabbit, or bovine extracts (in Figs. Z-5). Degradation of H20t-modified SOD during incubation with Fraction II increased with the length of incubation (Fig. 7A). In addition, as also demonstrated with less pure cell-free extracts above, degradation was also dependent upon the duration of pretreatment (exposure time) with H202 (Fig. 7B).

The activity responsible for the selective degradation of HpOz-modified SOD was purified, by (NH&SO., fractionation and gel filtration chromatography, to a 670-kDa proteinase complex (Table IV). This proteinase complex has previously been shown to catalyze the selective degradation of ‘OH- modified hemoglobin and albumin and we have suggested the trivial name macroxyproteinase (abbreviated from macro- oxy-proteinase) or M.O.P. (27, 28). Erythrocyte and reticulocyte M.O.P. exhibited more than 2000-fold greater specific activity (with comparable selectivity) in the degradation of HpOp-modified SOD than did cell lysates (Table IV). Further purification of M.O.P. by ion exchange chromatography (DEAE-Sepharose CL-6B) resulted in only a doubling of

FIG. 7. Degradation of HsOz-modified SOD by a Fraction II from human erythrocytes. [sH]SOD was exposed to 0, 1.5, or 30 mM Hz02 for up to 20 h (24 “C, pH 7.8), then extensively dialyzed. Treated and control samples were then incubated with erythrocyte Fraction II, purified by DEAE chromatography from human red cell extracts as described under “Experimental Procedures.” Incubations contained 43.3 fig of Fraction II protein and 3.3 pg of [3H]SOD protein (final volume of 0.13 ml), and were carried out for up to 5.5 h at 37 “C (pH 7.8). Panel A shows the effect of varying the length of incubation with Fraction II, following Hz02 (or “sham”) exposure for 1 h. Panel J3 demonstrates the effect of varying H202 exposure time on the degradation of SOD during a subsequent incubation with Fraction II for 5.5 h. Symbols used are as follows: 0, no HZ02 treatment (samples were incubated and dialyzed); A, exposure to 1.5 mM H,O,; 0, exposure to 30 mM H202. Proteolysis was measured as described under “Experimental Procedures,” and values shown are means of three independent determinations for which SE. was always less than 10%.

specific activity (compared with Table IV), and the enzyme complex was somewhat modified by the procedure as evidenced by increased activity against untreated [3H]SOD (data not shown). Electrophoresis (SDS-PAGE) of the 670-kDa M.O.P. complex revealed the presence of 8 polypeptides in the molecular mass range 21.5 to 35.7 kDa (data not shown).

Fragmentation of PHISOD as a Function of HzOz Concen- tration and Duration of Exposure-Direct fragmentation of proteins has been observed following exposure to certain oxidants (18-22, 25, 30-32, 34, 35). Direct fragmentation of SOD cannot explain the proteolysis data of Figs. 2-7 or Tables III and IV because all SOD samples were extensively dialyzed (against a membrane with a molecular size cut off of 6000- 8000) prior to incubation with cell-free extract, or partially purified fractions. Acid-soluble counts, without RBC extract or fraction incubation, were always measured to ensure that small fragments did not survive the dialysis step. Fragmen- tation without proteolysis also cannot explain the results of Table I, since fragmentation alone cannot generate free alanine.

Direct fragmentation of [3H]SOD by H202 was measured by an increase in acid-soluble counts without dialysis, and without incubation with red cell extract or fractions. Exposure of [3H]SOD to 1.5 mM HzOz resulted in very limited fragmentation during 1 or 2 h of incubation, but approximately 10% fragmentation occurred when incubation was extended to 20 h (Fig. 8). When [3H]SOD was exposed to 30 mM H202, fragmentation increased to approximately 35-40% after 20 h incubation. Direct fragmentation (following prolonged incubation) was also measured by both loss of the SOD monomer band in denaturing electrophoresis gels, and by acid-soluble counts, as a function of HzOz concentration (Fig. 9). Impor- tantly, electrophoresis revealed significantly greater SOD fragmentation (Fig. 9) than was evident from measurements of acid-soluble counts (Figs. 8 and 9). Since only fragments with molecular weights of approximately 5,000 daltons or less are soluble in trichloroacetic acid (as for peptides), these results indicate that many of the SOD fragments generated by prolonged exposure to high concentrations of H202 had


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/


TABLE IV Purification of the proteolytic activity selective for oxidatiuely modified superoxide dismutase: M 0.P

[‘HISOD was exposed to 1.5 mM HnOr (24 “C, pH 7.8) and extensively dialyzed. Untreated [‘HISOD was also incubated (60 min, 24 “C, pH 7.8) and dialyzed (see “Experimental Procedures”). Both untreated and HLOL-treated [‘HISOD were then incubated with RBC fractions (60 min. 37 “C, pH 7.8) and proteolysis was measured as described under “Experimental Procedures.” RBC lysates, cell-free extracts, and Fraction II were prepared from rabbit erythrocytes and reticulocytes as described under “Experimental Procedures.” Further purification of the (H,OI-modified) SOD degrading activity of Fraction II was achieved by ammonium sulfate precipitation (61). Both a O-40% (NH,)SOI precipitate and a 40-80% (NH.,)$SO, precipitate were prepared by adding solid (NH,),SO, to Fraction II and stirring for 30 min at 4 “C. Precipitated proteins were pelleted by centrifugation at 3000 X g for 20 min, and the precipitates were redissolved and dialyzed (3500 cut off membrane) to remove salt. The redissolved O-40% (NH,),SO, precipitate exhibited little proteolytic activity and was discarded. The redissolved 4080% (NH,),SO, precipitate retained the activity of Fraction II and was further purified by gel filtration chromatography on a 1.6 x 100~cm Sephacryl S-300 (Pharmacia LKB Biotechnology Inc.) column. The Sephacryl resin was pre- equilibrated with a buffer containing 20 mM Tris-HCI (pH 7.8), 20 mM KCl, 1 mM MgOAc, and 0.5 mM DTT. Fractions (2.5 ml) were eluted (15 ml/h) with the same buffer and screened for proteolytic activity with HlO,- modified [ ‘HISOD. The activity eluted in a single peak of 670 kDa which we have named macroxyproteinase or M.O.P. (27, 28). Values are means of three independent determinations for which S.E. was always less than 10%. Similar results were also obtained with RBC fractions, and M.O.P., from bovine and human erythrocytes (data not shown).

Superoxide dismutase degradation

RRC fraction Erythrocyte fractions

Untreated HIOX-treated SOD SOD

Reticulocyte fractions

Untreated H,Os-treated SOD SOD

pg WI1 degraded. mg protan-‘. h-’

Lysate 0.003 0.039 0 0.012 Extract 0.001 0.020 0 0.010 Fraction II 0.19 1.38 0.26 2.68 40-80% (NH,),SO, fraction 0.05 1.47 0.25 2.82 Macroxyproteinase 5.5 112.0 0 29.2

Incubotlon Time (hrs)

FIG. 8. Production of low molecular weight fragments of SOD as a function of H202 exposure time. [‘HISOD was exposed (see “Experimental Procedures”) to 1.5 (A) or 30 mM (0) HLOL, or was used as an unexposed control (0). Samples were incubated (with no further additions) for up to 20 h at 24 “C (pH 7.8). Catalase (5 fig) was added after 0, 0.5, 1, 2, or 20 h to remove any residual H?O,, but samples were not dialyzed. Fragmentation was determined as the percent of total (initially acid precipitable) counts which were con- verted to acid-soluble (<5000 daltons) counts. Note that SOD samples were not incubated with RBC extracts or proteolytic fractions. Values are means f S.E. of three independent determinations.

molecular sizes between 5,000 and 16,000 (the size of the SOD monomer).

While conditions were designed to rule out the possibility of autodegradation by contaminating proteases in our preparations (i.e. extensive dialysis of samples) we further tested this possibility with additional experiments. Untreated and 30 mM HZOZ-treated [“HISOD (1 h) were extensively dialyzed and then incubated for 20 h at 37 “C! in the presence of equimolar concentrations of untreated [“HISOD (as a potential source of contaminating proteases). Following incubation, acid-soluble counts were measured as described under “Ex- perimental Procedures.” We observed 0.29% conversion to acid-soluble counts for (dialyzed) untreated [“HISOD and

A A--- - . . 0 1.5 4.5 9 15 30

H,O, CmM)

FIG. 9. Differential SOD fragmentation profiles revealed by electrophoresis and acid-soluble counts as a function of H202 concentration. [“HISOD was exposed to various concentrations of HIOZ (or was used as an unexposed control) during incubation for 20 h at 24 “C (pH 7.8). Incubations were terminated upon the addition of catalase (5 pg), but samples were not dialyzed. Fragmen- tation was determined both by loss of the (16 kDa) SOD monomer band in denaturing (12.5%) SDS-PAGE, and by measurement of the production of acid-soluble counts. Panel A shows a representative SDS-PAGE of [“HISOD after treatment with various concentrations of H,O,. Panel B shows SOD fragmentation both by quantification of the SOD SDS-PAGE band (O), and by measurement of the production of acid-soluble counts (0) (liquid scintillation). SDS- PAGE bands were visualized with silver stain, and quantified by scanning densitometry with computerized integration as previously described (19). Values in panel B are means of three independent determinations for which S.E. was always less than 10%.

0.73% for (dialyzed) 30 mM H20L-treated [“HISOD. This low generation of acid-soluble counts could actually be explained by exchange reactions of [“HI with the surrounding water.

0 10 20 30 H,O, (mt4)


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/


Under similar conditions the conversion to acid-soluble counts of undialyzed samples were 0.8 and 22.9% for untreated and 30 mM HzOz-treated [3H]SOD, respectively. We also performed the same experiment utilizing [3H]casein (as a substrate) in the presence of equimolar concentrations of untreated [3H]SOD (as a potential source of contaminating proteases). Only approximately 0.98% conversion to acid- soluble counts was measured despite the fact that casein is an exceptionally good proteolytic substrate. It appears certain that fragmentation of SOD as reported is not the result of contaminating proteases in our preparations.

Red cell SOD is continuously exposed to 0, and H202 in uivo via hemoglobin autooxidation and the metabolism of a wide variety of toxicants (l-7). Thus, SOD fragmentation may also be of distinct biological significance. Accumulation of protein fragments may in fact present a substantial toxi- cological challenge to RBC and other cells. Such fragments would be expected to undergo rapid degradation by intracellular peptidases (18-22, 25, 30-32, 34, 35), although this proposal has not yet been tested.

strong correlation between ’ OH-induced increases in hydrophobicity and proteolytic susceptibility (18-22, 25-32), we suggest that Hz02 may cause a partial denaturation or unfolding in the active site region of SOD. Such denaturation may reveal previously shielded hydrophobic amino acid residues which may be preferred proteolytic substrates. In the absence of proteolytic enzymes in uitro, HzOB-modified SOD undergoes further reactions to produce protein fragments. It appears probable that H202 (or 0;) inactivation of SOD in uivo produces both modified (but intact) proteolytic substrates, and direct fragmentation products. Such fragments (which could have untoward biological effects) would be similar to true peptides (except for their modified amino and carboxyl termini) and may be rapidly degraded by intracellular peptidases.

DISCUSSION

Our results indicate that oxidative inactivation of red cell SOD by its product H202 generates a modified protein which is recognized and selectively degraded by an intracellular proteolytic pathway. To our knowledge, this is the first report in which SOD inactivation by H,O, and 0; in intact cells has been correlated with the inactivation of the purified enzyme in uitro. Previous studies have examined the inactivation of the purified enzyme alone (9-17). We report that increased proteolysis accompanies SOD inactivation in intact RBC, and purified SOD exhibits increased proteolytic susceptibility (to red cell proteases) following inactivation by H202 in vitro. On a molar basis we find that H202 is approximately 2-3 times more efficient in inactivating SOD in vitro than in intact erythrocytes. The higher “resistance” of SOD to Hz02 in RBC is presumably due to the presence of catalase and glutathione peroxidase. This interpretation is strengthened by the finding that addition of KCN (which inhibits catalase) greatly enhances proteolysis (23).

The selective degradation of HzOz-modified SOD in red cell extracts is now seen to be catalyzed by an ATP-independent proteolytic pathway (in fact, ATP-Mg2+ was slightly inhibi- tory). This pathway is clearly distinct from the ATP-ubiquitin proteolytic system reported in reticulocytes (41-43), although both may share a common pool of peptidases. The proteolytic pathway for HzOp-modified SOD was strongly inhibited by sulfhydryl reagents, serine reagents, and transition metal chelators; suggesting the involvement of sulfhydryl-, serine-, and metalloproteinases, proteases, and peptidases. Hemin appears to inhibit many forms of proteolysis in red cells (by an unknown mechanism) and was also effective in this study. Importantly, leupeptin did not inhibit the degradation of HzOz-modified SOD, implying that lysosomal and calcium- activated thiol proteases were not involved. Essentially the same inhibition profile for red cell extracts was previously reported for proteolysis of albumin, hemoglobin, catalase, and SOD following oxidative denaturation by ‘OH (18-22, 25- 28). Thus, our previous work with artificial oxygen radical- generating systems and “foreign” proteins apparently serves as a good model for the proteolysis of naturally occurring oxidized proteins in uiuo.

Interestingly, SOD inactivation and protein degradation was greater when intact RBC were exposed to a continuous flux of 0; and H202, generated by xanthine + xanthine oxidase, than when the cells were exposed to a bolus of H202; even though the maximum production of 0; + Hz02 by xanthine oxidase (1 mM) was only 7% of the highest H202 bolus addition (15 mM). We suggest that these results reflect the much greater affinity of SOD for its substrate 0; (as well as other anions, such as HO;), than for its product H202 (8, 14). It appears probable that SOD inactivation in uiuo may be more efficient when HzOz is generated at the enzyme’s active site by 0; dismutation, than when cells are exposed to a bolus of H202 (much of which could be removed by catalase and glutathione peroxidase). Alternatively, 0; and HO; (the conjugate base of H202) may act synergistically to generate an oxidizing species, at the active site copper, which may attack one of the liganding imidazoles.

In this article we also report the degradation of H202- modified SOD by a 670-kDa red cell proteinase complex which was initially purified on the basis of its ability to degrade ‘OH-modified hemoglobin (27, 28). We have tentatively named this enzyme macroxyproteinase (macro-oxy-proteinase) or M.O.P. (26-28); macro- because of its size, -oxy- because of its clear preference for oxidatively moditied substrates, and -proteinase as a generic suffix because it contains sulfhydryl-, serine-, and metal-dependent activities. M.O.P., which has maximal activity at pH 7.8, may be synonymous with or at least closely related to neutral/alkaline proteinases of 600-700 kDa size which have been isolated from a variety of tissues (44-54). These large proteinase complexes appear to possess several catalytic activities with artificial substrates in vitro and a variety of trivial names have been suggested (44-54), although no clear cellular function has previously been found. The degradation of oxidatively modified proteins provides the first clear intracellular role for this 670-kDa proteinase complex, and we suggest that the name macroxyproteinase (M.O.P.) provides structural and functional infor- mation without prejudice to the apparently conflicting reports of inhibitor profiles (44-54).

Both loss of SOD activity and modified active site copper binding appear to precede proteolytic recognition and degradation. SOD inactivation by H202 has been shown to involve oxidative modification of a histidine residue which is important in the binding of the copper moiety to the active site of the enzyme (8-10,12, 13). We suggest that the HzO2-dependent structural changes in the active site region of SOD represent a conformational change which is susceptible to pro- _ teolysis. Since our work with other proteins has revealed a _~ -- -

A 650-kDa proteinase complex was isolated from rodent liver by Rivett (55, 56) on the basis of its ability to degrade glutamine synthetase, following modification by mixed function oxidase systems. We have previously proposed (for both RBC and Escherichiu coli (18-22, 25-28, 31, 32)) that increased hydrophobicity is the key to selective degradation of nroteins which have been modified bv active oxygen species


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/


such as ‘OH and H202. Cervera and Levine (57), building upon earlier work (5%61), have also proposed that the selective degradation of glutamine synthetase (in bacteria) is promoted by a mixed-function oxidase-dependent increase in hydrophobicity. It, thus, seems reasonable to propose that Rivett’s (55,56) enzyme and M.O.P. may actually be one and the same proteinase complex, which may have broad proteolytic activity against a variety of oxidatively modified substrates.

Red blood cells experience a continuous flux of 0; and H202 due to hemoglobin autooxidation (l-4). Various drugs and environmental agents add to the oxidative stress to which red cells are exposed (5-7). Mature erythrocytes have no capacity for de nouo protein synthesis, and cannot replace oxidatively damaged SOD. We propose that (during oxidative stress) M.O.P. catalyzes the recognition and initial degradation of HPOz-modified SOD in erythrocytes, thus preventing the accumulation of SOD aggregates and diminishing the formation of SOD fragments. In this scheme, M.O.P. would constitute a key component of “Secondary Antioxidant De- fenses” or “Repair Systems” (18, 25). Reticulocytes are capa- ble of de nouo protein synthesis and may even mount an oxidative stress response, involving the overexpression of SOD, to oxidants such as H202; as do other cells (63-65). Thus, it is possible that M.O.P. may participate in a coordi- nated system for SOD synthesis and degradation in reticulocytes (and other cells), in which all aspects of SOD turnover are regulated by 0; (or H202).

Acknowledgment-We wish to thank Dr. Sandra Jewett (California State University, Northridge) for suggesting and demonstrating the bathocuproine assay for SOD active site copper binding, and for sharing her unpublished work on this assay.

REFERENCES 1. Chin, D., Lubin, B., and Shohet, S. B. (1982) in Free Radicals in Biology

(Pryor, W. A., ed) Vol. 5, p. 115-160, Academic Press, Orlando, FL 2. Heppel, R. P., Eaton, J. W., Balasingan, M., and Steinberg, M. H. (1982)

J. Clin. Invest. 70, 1253-1259 3. Misra, H. P., and Fridovich, I. (1972) J. Biol. Chem. 2.47, 6960-6962 4. Wever, R., Oudega. B., and Van Gelder, B. F. (1973) Biochim. BioDhvs.

Act&30i, 75-78 . -

5. Lal, A. K., Ansari, N. H., Awasthi, Y. C., Snyder, L. M., Fortier, N. L., and Srivastava. S. K. (1980) J. Lab. Clin. Med. 94.536-552

6. Cohen, G., and Ho&stein, P. (1964) Biochemi&y 3,895-900 7. Cohen, G., and Hochstein, P. (1965) J. Phrmacol. Exp. Ther. 147, 139-

143 8. Bly7$-:86M., and Borders, C. L., Jr. (1983) Arch. Biochem. B[ophys. 224,

9. Borders, C. L., and Fridovich, I., Jr. (1985) Arch. Biochem. Biophys. 241, 472-476

10. Hodgson, E. K., and Fridovich, 1. (1975) Biochemistry 14,5294-5299 11. Rotilio. G.. Mornureo. L.. Calabrese. L.. and Mondovi. B. (1973) Biochim.

Biophys. Acta30&‘22&235 12. Bray, R. C., Cockle, S. A., Fielden, E. M., Roberts, P. B., Rotilio, G., and

Calabrese, L. (1974) Biochem. J. 139,43-48 13. Forman, H. J., Evans, H. J., Hill, R. L., and Fridovich, I. (1973) Biochem-

istry 12,823-827 14. Cndd, A., and Fridovich, I. (1982) J. Biol. &em. 257,11443-11447

15.

16. 17.

18. 19. 20.

21.

Fuchs, H. J. R., and Borders, C. L., Jr. (1983) Biochm. Biophys. Res. Commun. 116,1107-1113

Hodgson, E. K., and Fridovich, I. (1975) Biochemistry 14,5294-5299 Jewett, S. L., Cushing, S., Gillespie, F., Smith, D., and Parks, S. (1989)

Eur. J. Biochem. 180,569-575 Davies, K. J. A. (1986) J. Free Radicals Biol. Med. 2, 155-173 Davies, K. J. A. (1987) J. Biol. Chem. 262,9895-9901 Davies, K. J. A., Delsignore, M. E., and Lin, S. W. (1987) J. Biol. Chem.

262,9902-9907 Dyi%; K. J. A., and Delsignore, M. E. (1987) J. Biol. Chem. 262, 9908-

22. Davies, K. J. A., Lin, S. W., and Pacifici, R. E. (1987) J. Biol. Chem. 262, 9914-9920

23. Davies, K. J. A., and Goldberg, A. L. (1987) J. Biol. &em. 262,8220-8226 24. Davies, K. J. A., and Goldberg, A. L. (1987) J. Biol. Chem. 262,8227-8234 25. Davies, K. J. A. (1988) in Cellular Antioxidant Defense Mechanisms (Chow,

C. K., ed) Vol. 2, pp. 25-67, CRC Press, Boca Raton, FL 26. Sale, D. C., Lin, S. W., Pacifici, R. E., and Davies, K. J. A. (1988) Free

Radical Biol. Med. 5, 335-339 27. Pacifici, R. E., and Davies, K. J. A. (1988) FASEB J. 2, 4135 (abstr.) 28. Pacifici, R. E., Salo, D. C., and Davies, K. J. A. (1989) Free Radical Biol.

Med. 7,521-536 29. Taylor, A., and Davies, K. J. A. (1987) Free Radical Biol. Med. 3, 371-377 30. Marcillat, O., Zhang, Y., Lin, S. W., and Davies, K. J. A. (1988) Biochem.

J. 254,677-683 31. Davies, K. J. A., and Lin, S. W. (1988) Free Radical Biol. Med. 6, 215-223 32. Davies, K. J. A., and Lin, S. W. (1988) Free Radical Biol. Med. 6, 225-236 33. Stadtman, E. R. (1986) Trends Biochem. Sci. ll, ll-12 34. W;:ff3S. P., Garner, A., and Dean, R. T. (1986) Trends Biochm. Sci. 11,

35. Wolff, S. P., and Dean, R. T. (1986) Rio&em. J. 234,399-403 36. Aebi, H. (1984) Methods Enrymol. 105, 121-126 37. O’Brien, P. J. (1984) Methods Enzymol. 106,370-378 38. Ciechanover, A., Hod, Y., and Hershko, A. (1978) Biochem. Biophys. Res.

Commun. 81, 1100-1105 39. McCord, J. M., and Fridovich, I. (1969) J. Biol. Chn. 244,60$9-6055 40. Bo;;;sl,,t, Martmo, E., and Stoppam, A. 0. M. (1977) Anal. &o&em. 80,

41. GoiMp A. L., and St. John, A. C. (1976) Annu. Rev. Biochem. 45, 747-

42. EtIi;gerr_J8D., and Goldberg, A. L. (1977) Proc. N&l. Acad. Sci. U. S. A.

43. Hersdkq, A., and Ciechanover, A. (1982) Anna Reu. Biochm. 51,335-364 44. De3~;;tmo, G. N., and Goldberg, A. L. (1979) J. Brol. Chem. 254, 3712-

45. Wilk, S., and Orlowski, M. (1980) J. Neurochem. 35,1172-1182 46. Wilk, S., and Orlowski, M. (1983) J. Neurochem. 40,842-849 47. Dald$nann, B., Kuehn, L., and Reinauer, H. (1983) FEBS Lett. 160, 243-

48. Ishiura, S., Sane, M., Kamakura, K., and Sugita, H. (1985) FEBS L&t. 189,119-123

49. Ra;54;., and Harris, H. (1985) Proc. Natl. Acad. Sci. U. S. A. 82, 7545-

50. M;v;$8gM. J., and DeMartino, G. N. (1986) Biochim. Biophys. Acta 873,

51. Waxman, L., Fagan, J. M., and Goldberg, A. L. (1987) J. Biol. Chem. 262, 2451-2457

52. H”s”3p R., Pratt, G., and Rechsteiner, M. (1987) J. Biol. Chem. 262,8303-

53. Falkenbur I,., and B

P. E., Haas, C., Kloetsel, P.-M., Niedel, B., Kopp, F., Kuehn, ahlmann, B. (1988) Nature 331, 190-192

54. Arrigo, A. P., Tanaka, K., Goldberg, A. L., and Welch, W. J. (1988) Nature 331,192-194

55. Rivett, A. J. (1985) J. Biol. &em. 260,300-305 56. Rivett, A. J. (1985) J. Biol. Chem. 260, 12600-12606 57. Cervera, J., and Levine, R. L. (1988) FASEB J. 2, 2591-2595 58. Levine, R. L., Oliver, C. N., Fulks, R. M., and Stadtman, E. R. (1981) Proc.

Natl. Acad. Sci. U. S. A. 78,2120-2124 59. Fucci, L., Oliver, C. N., Coon, M. J., and Stadtman, E. R. (1983) Proc. N&l.

Acad. Sci. U. S. A. 80.1521-1525 60. Levine, R. L. (1983) J. Biol. Chem. 258, 11823-11827 61. Levine, R. L. (1983) J. Biol. Chem. 258, 1182811833 62. Green, A. A., and Hughes, W. L. (1955) Methods Enzymol. 1,67-90

64. d 63. S’ostrom K., and Crapo, J. D. (1983) Lab. Inuest. 48, 68-79

ass, M A, and Massaro, D. (1988) J. Biol. &em. 263, 776-781 .' 65. Hass, M. A., Iqbal, J., Clerch, L. B., Frank, L., and Massaro, D. (1989) J.

Clin. Invest. 83, 1241-1246 66. Salo, D. C., Lin, S. W., Pacifici, R. E., and Davies, K. J. A. (1988) FASEB

J. 2, 4136 (abstr.)


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

D C Salo, R E Pacifici, S W Lin, C Giulivi and K J Daviesoxidative modification and inactivation.

Superoxide dismutase undergoes proteolysis and fragmentation following

1990, 265:11919-11927.J. Biol. Chem.

http://www.jbc.org/content/265/20/11919Access the most updated version of this article at

Alerts:

When a correction for this article is posted•

When this article is cited•

to choose from all of JBC's e-mail alertsClick here

http://www.jbc.org/content/265/20/11919.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/content/265/20/11919

http://www.jbc.org/cgi/alerts?alertType=citedby&addAlert=cited_by&cited_by_criteria_resid=jbc;265/20/11919&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/265/20/11919

http://www.jbc.org/cgi/alerts?alertType=correction&addAlert=correction&correction_criteria_value=265/20/11919&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/265/20/11919

http://www.jbc.org/cgi/alerts/etoc

http://www.jbc.org/content/265/20/11919.full.html#ref-list-1

http://www.jbc.org/

superoxide dismutase undergoes proteolysis and ...superoxide dismutase undergoes proteolysis and...

Documents