IMephrology Dialysis Transplantation

Nephrol Dial Transplant (1996) 11 [Suppl 5]: 34-40 IMephrology Dialysis Transplantation Oxidative stress caused by glycation of Cu,Zn-superoxide dismutase and its effects on intracellular components Junichi Fujii, Theingi Myint, Ayako Okado, Hideaki Kaneto and Naoyuki Taniguchi Department of Biochemistry, Osaka University Medical School, 2-2 Yamadaoka, Suita, Osaka 565, Japan Abstract. It is now evident that the redox state of the cell is a pivotal determinant of the fate of cells. Extensive production of reactive oxygen species (ROI) causes necrotic cell death. Even transient or localized production of ROI may mediate a signal for apoptotic cell death, whereas small amounts of ROI function as an intracellular messenger of some growth stimulants. Accumulating evidence supports the concept that decreases in Cu,Zn-superoxide dismutase (SOD) activity causes apoptotic cell death in neuronal cells. Our data using mutant Cu,Zn-SOD related to familial amyotrophic lateral sclerosis (FALS) suggest that glycation itself and ROI produced from the glycated proteins are involved in many diseases, including diabetic complications. Glycation of important cellular components, including lipid, DNA and proteins, induces dysfunction of these components. Mutant proteins in patients with various hereditary diseases would be destabilized by the glycation reaction, as shown in the case of mutant Cu,Zn-SODs, thereby hyperglycaemic conditions would trigger the onset of some hereditary diseases such as FALS and Alzheimer's disease. Glycation, particularly of antioxidative enzymes, would enhance production of ROI, resulting in oxidative damage to the cells. Key words: aldo-keto reductase; amyotrophic lateral sclerosis; reactive oxygen intermediates; superoxide dismutase Introduction Glycation is thought to occur during normal ageing with accelerated rates in diabetes mellitus and to be involved in structural and functional changes of proteins. It has been suggested that the glycation reaction is involved in the pathogenesis of diabetic complications [1]. Glucose- and fructose-derived Correspondence and offprint requests to: Naoyuki Taniguchi, Department of Biochemistry, Osaka University Medical School, 2-2 Yamadaoka, Suita, Osaka 565, Japan. 1996 European Renal Association-European Dialysis and Transplant Association metabolites are responsible for protein modification and cross-linking of proteins. The cross-linking of long-lived proteins such as collagen and lens crystallin correlates with ageing and diabetes. Furthermore, glycation alters the activity of some enzymes such as Cu,Zn-superoxide dismutase (SOD) [2,3], carbonic anhydrase [4], alcohol dehydrogenase [5], aldehyde reductase [6] and Na,K-ATPase [7]. Several groups have reported that Schiff base or Amadori products generate reactive oxygen intermediates (ROI) [8,9], which are assumed to be involved in many diseases, including cancer, diabetes and neuronal dysfunction. Involvement of divalent cations in the production of reactive oxygen species by the glycation reaction has been suggested. Defence systems against increase in glycation Our body has defence systems against various stimuli, and that against glycation reaction seems to exist in most living organisms. We have purified a rat liver enzyme that catalyses the NADPH-dependent reduction of 3-deoxyglucosone (3-DG) [10], a major intermediate in the Maillard reaction and a potent cross-linker responsible for the polymerization of proteins. Comparison of the amino acid sequences of nine peptides obtained from the rat 3-DG-reducing enzyme by lysyl-endopeptidase digestion with the amino acid sequence of human aldehyde reductase strongly suggested that the purified enzyme was rat aldehyde reductase. We cloned the cdna encoding aldehyde reductase from a rat kidney cdna library using a human aldehyde reductase cdna fragment, amplified by polymerase chain reaction, as a probe. All nine peptides identified in the purified rat 3-DG-reducing enzyme were found in the amino acid sequence deduced from the rat aldehyde reductase cdna. Moreover, cell extract from COS-1 cells transfected with the rat aldehyde reductase cdna exhibited NADPH-dependent 3-DG-reducing activity and cross-reacted with antiserum raised against the purified rat 3-DG-reducing enzyme. All these data indicate clearly that the 3-DG-reducing enzyme is identical to aldehyde reductase in rat liver. Northern

Glycation of Cu,Zn-SOD 35 blot analysis of total RNA from a variety of rat tissues showed fairly high levels of expression of aldehyde reductase mrna. This suggests that aldehyde reductase present in the body would detoxify 3-DG when it is formed through the Maillard reaction in vivo. Aldehyde reductase purified from normal rat liver was found to be partially glycated as judged by binding to a boronate column and reactivity to anti-hexitol lysine IgG [6]. Sites of in vivo glycation of rat liver aldehyde reductase were identified by sequencing of digested peptides labelled with NaB[ 3 H] 4 and by mass spectrometry. The major glycated sites were found to be lysines 67, 84 and 140. The glycated enzyme had a lower catalytic efficiency than the non-glycated form. In streptozotocin-induced diabetic rats the glycated form was significantly increased in the kidneys. Because the enzyme plays a role in detoxifying 3-DG formed through the Maillard reaction in vivo, glycation of aldehyde reductase and reduction of its activity may result in metabolic imbalance and accelerate the glycation reaction under diabetic conditions. Aldose reductase, an enzyme constituting the polyol pathway together with sorbitol dehydrogenase, is in the aldo-keto reductase gene family and has significant homology with aldehyde reductase. This enzyme also can reduce 3-DG in the same mechanism as aldehyde reductase. Since accumulation of sorbitol is thought to play a role in the pathogenesis of diabetic complications in some organs such as kidney, peripheral nerve and eye, extensive studies on aldose reductase have been carried out to elucidate the causal connection between the activity of this enzyme and diabetic complications [11]. Various inhibitors of this enzyme have been developed as a therapy for diabetes mellitus. Several enzymes in this gene family were found to play an important role in the resistance of cells to the cytotoxicity of compounds with an aldehyde moiety such as tripeptidyl aldehyde, trioses and methotrexate. We examined the concentrations of protein and mrna of aldose reductase and aldehyde reductase to see whether or not the reducing enzyme was essential for the detoxification of aldehyde compounds including 3-DG in hepatocarcinogenesis [12]. While the amount of aldehyde reductase did not change, aldose reductase gene induction was observed in hepatoma tissues and cultured hepatoma cell lines. We also found that aldose reductase gene expression was induced during hepatitis in the Long-Evans rat with cinnamon-like coat colour (LEC) rats which spontaneously develop hepatoma [13]. Thus this enzyme seems to function in the detoxification of naturally occurring carbonyl compounds and may have a role in the immortality of malignant cells. Oxidative stress caused by glycation of Cu,Zn-SOD Glycation ofcu,zn-sod in vivo Superoxide dismutase (SOD) is an anti-oxidative enzyme which scavenges superoxide radical in the cytoplasm of most mammalian cells [14]. Among three SOD isozymes, Cu,Zn-SOD present in the cytoplasm is a dimeric protein with copper and zinc ions in its catalytic centre. Human erythrocytes contain glycated and non-glycated Cu,Zn-SOD, which can be separated by boronate affinity chromatography [2]. The percentage of the glycated form is significantly increased in the erythrocytes of patients with diabetes compared with normal erythrocytes. The non-glycated form of Cu,Zn-SOD, which was washed through the boronate column, was glycated in vitro upon exposure to D-glucose. Incorporation of D-glucose into the protein was observed, and with the increase in glycation the enzymatic activity decreased, indicating that the glycation of the enzyme led to a low active form. We also examined the amount and nature of Cu,Zn-SOD in insulin-dependent diabetic children [15]. Glycated Cu,Zn-SOD, which binds to a boronate affinity column, was measured by the enzyme-linked immunosorbent assay. The percentage of the glycated form in 25 insulin-dependent diabetic children was significantly greater than that in the normal controls, while its specific activity was significantly lower. These data indicate that glycated and less active Cu,Zn-SOD is increased in erythrocytes of patients with insulindependent diabetes mellitus. Identification of specific residues in glycated Cu,Zn-SOD Since glycated Cu,Zn-SOD exhibits less activity than the non-glycated form, the sites of the glycation reaction were identified in vitro [3]. Purified Cu.Zn- SOD from human erythrocytes generally comprises both glycated and non-glycated forms. The nonglycated Cu,Zn-SOD was isolated by boronate affinity chromatography from glycated protein. Incubation of the non-glycated SOD with D-[6-3 H]glucose in vitro resulted in the gradual accumulation of radioactivity in the enzyme protein, and Schiff base adducts were trapped by NaBH4 reduction. The sites of glycation of Cu,Zn-SOD were identified by amino acid analysis after reverse-phase HPLC of the trypsin-treated peptides. The lysine residues Lys3, Lys9, Lys30, Lys36, Lysl22 and Lysl28 were found to be glycated. Three of the glycated sites lie in Lys-Gly, two in Lys-Ala and one in Lys-Val. The inactivation of the SOD on the glycation is due mainly to the glycation of Lysl22 and Lysl28, which are supposed to be located in an active site ligand-binding loop. The remaining five sites are relatively inactive in the formation of Amadori adducts. The effects of glycation reaction on mutant Cu,Zn-SODs at Lysl22 and Lysl28 are currently under investigation. Fragmentation of glycated Cu,Zn-SOD We further characterized glycated Cu,Zn-SOD and found that the glycated enzyme was degraded by the glycation reaction [16], as briefly described below. Incubation of human Cu,Zn-SOD with 0.1 M glucose at 37 C resulted in a time-dependent decrease

36 J. Fujii et al. in the amount of intact enzyme. Sodium dodecylsulphate-polyacrylamide gel electrophoresis showed a gradual decrease in the intensity of the original band with a molecular mass of ~20 kda and the simultaneous appearance of a high-molecular-weight fragment with a distinct 15 kda band and a small molecular mass fragment of ~5 kda. Thus this cleavage occurred in a rather site-specific manner during the initial stage of the Maillard reaction, and then random fragmentation was observed after 7 days. In order to identify what causes the fragmentation of SOD, ESR spectra of the radical species produced from the glycated Cu,Zn-SOD were measured after incubation with 0.1 M glucose for 3 days using 5,5-dimethyl-l-pyrroline-iV-oxide (DMPO) as a spin trapping agent. Glycation of Cu,Zn-SOD generated a hydroxyl radical, which was detected by ESR as a DMPO-OH adduct. No signals were observed after addition of EDTA, which suggests involvement of metal ions in this reaction. Addition of catalase also prevented the appearance of the hydroxyl radical signals, suggesting that H2O2 is also involved in this reaction. In the first step, Cu,Zn-SOD was cleaved at a peptide bond between Pro62 and His63, as judged by amino acid analysis and sequencing of fragment peptides. The same fragmentations were observed upon exposure of the enzyme to an H 2 O 2 bolus. Catalase completely blocked only the second step of the fragmentation process, whereas EDTA blocked both steps. Incubation with glucose resulted in the time-dependent release of Cu 2+ from Cu,Zn-SOD. The release of Cu 2+ then probably participated in a Fenton-type reaction to produce the hydroxyl radical which may cause the non-specific fragmentation. Evidence that EDTA abolished only the second step of fragmentation induced by an H2O2 bolus supports this mechanism. These data suggest that a superoxide anion is generated from the glycation reaction and then hydrogen peroxide is formed by a dismutation reaction of Cu,Zn-SOD, resulting in conversion to a hydroxyl radical which causes the first site-specific cleavage. This brings about inactivation of the enzyme and release of the copper ions. Free copper ions would then mediate the Fenton reaction and the resultant hydroxyl radical would cause further non-specific cleavage of the peptides. We also found that ceruloplasmin, a copper-containing plasma protein, also led to fragmentation by glycation in an ROI-mediated manner [17]. This ROI-mediated site-specific cleavage may occur in any glycated proteins containing transition metal ions. DNA fragmentation by glycated Cu, Zn-SOD The reactive oxygen species damages many biological molecules, including DNA. Cleavage of purified DNA is known to be induced by direct treatment with metals plus hydrogen peroxide. DNA strand breaks and fragmentation induced by alloxan in pancreatic islet cells seem to play an important role in the development of diabetes. We therefore investigated the effects of glycated Cu,Zn-SOD on cloned DNA fragments and nuclear DNA [18]. Incubation of a radiolabelled DNA fragment with glucose, bovine serum albumin, Cu,Zn-SOD and their combination for 5 days was followed by gel electrophoresis. The DNA cleavage occurred only in samples incubated with glycated Cu,Zn-SOD and the cleavage occurred at the deoxyribose-phosphate backbone. The extent of DNA cleavage increased with time and with increasing concentrations of Cu,Zn-SOD and glucose. Incubation with fructose instead of glucose resulted in more rapid and marked DNA cleavage. All these data are consistent with the idea that DNA cleavage was induced by glycated SOD. We also examined whether glycation of Cu,Zn-SOD actually damages DNA in nuclei. Nuclei isolated from cultured HIT cells and K562 cells were treated with 1 mg/ml Cu,Zn-SOD pre-incubated with 0.1 M glucose in phosphate-buffered saline for 10 days. Nuclear DNA was found to be cleaved by incubation with copper and hydrogen peroxide. Cleavage of nuclear DNA was also observed when nuclei were incubated with Cu,Zn-SOD with glucose but not SOD or glucose alone. Several hydroxyl radical scavenging agents could prevent the DNA cleavages. This DNA cleavage was inhibited by catalase or metal chelating reagents such as EDTA, which clearly indicates that the reactive oxygen species, most probably a hydroxyl radical, is produced from glycated SOD and participates in cleavage of the nuclear DNA in isolated nuclei also. Hydroxyl radicals are known to modify DNA and cause mutation of genes. A major DNA adduct with a hydroxyl radical is 8-hydroxyldeoxyguanosine. To evaluate a role of the reactive oxygen species in the DNA modification, we measured 8-hydroxyldeoxyguanosine in DNA which had been incubated with glycated SOD. Although a small amount of 8-hydroxyldeoxyguanosine was detected in control DNA, as previously reported, it was not increased after incubation without SOD or glucose. With the addition of glycated SOD, the amount of 8-hydroxyldeoxyguanosine was substantially increased up to eight times that of the control after 5 days of incubation. The time course of DNA cleavage and formation of 8-hydroxyldeoxyguanosine indicates that the cleavage of the phosphodiester bond of DNA occurs before base modification with the reactive oxygen species. Hypothetical role of glycation in amyotrophic lateral sclerosis Mutations in Cu,Zn-SOD gene (SOD 1) andfals Since defects in the gene encoding Cu,Zn-SOD (SOD 1) were identified in familial amyotrophic lateral sclerosis (FALS), a motoneuron disease, >40 mutations linked to FALS have been found in 20-40% of FALS by many groups, including us [19-22]. Some of these mutations were assumed to affect the subunit interaction or folding of the enzyme [23]. Several mutant Cu,Zn-SOD

Glycation of Cu,Zn-SOD 37 have been expressed in mammalian cells and yeast, and organic solvents, we examined the nature of the mutant their instabilities have been demonstrated in cells Cu.Zn-SODs in comparison to the wild-type enzyme. [24,25]. It is not clear, however, whether these mutant Inactivation of enzymes is prominent in all samples enzymes have actually decreased SOD activity by incubated at 37 or 42 C for 24 h in phosphate-buffered reducing protein stability or it increased the saline in comparison with those incubated at 4 C, susceptibility of the enzymes to proteases due to especially the G37R and G85R mutants. The stability misfolding. The mechanism by which FALS is caused of mutants G41D and G41S decreased markedly at seemed to be more complex because both suppression 42 C. The activities of G85R were decreased and overexpression of Cu,Zn-SOD activity induced dramatically at this temperature. We also examined the neuronal cell death [26-29], although SOD activity sensitivity of the SOD activity of three mutant enzymes actually decreased by 40-60% in erythrocytes of FALS to the metal-chelating agent EDTA. Although the patients [21,23]. Several groups have hypothesized that activities of the wild-type and I113T SODs were not the gain of function in these mutant enzymes causes the changed after 8 h incubation with 5 mm EDTA, G85R disease based on the data from transgenic mice and G41D were inactivated by this treatment to overproducing mutant Cu,Zn-SOD. Because SOD different extents. Thus some of the mutant SODs found catalyses only the reaction of 2O2 " + 2H -» H2O2 + O2, in FALS seemed to have a decreased affinity for copper the existence of catalase or glutathione peroxidase is ions, suggesting that copper can be released from such essential to detoxify the resultant H2O2- Imbalance mutant SODs and enhance the Fenton reaction. between SOD and the latter enzymes, therefore, Since Cu,Zn-SOD is known to be inhibited by increases the risk of production of hydroxyl radical, the most toxic oxidant, by either the Fenton-type reaction hydrogen peroxide [30], which is the product of the or catalysis by Cu,Zn-SOD itself [30-32]. Another enzymatic reaction of SOD with superoxide, the effects possibility is that the Cu 2+ ion may be released from of hydrogen peroxide on the enzyme activity were unstable mutant enzyme and exhibit toxicity by examined. When wild-type Cu,Zn-SOD was incubated accelerating the Fenton reaction. with 1 mm hydrogen peroxide for various lengths of time, the SOD activity decreased in a time-dependent manner. We also determined copper ions released from Production of Cu, Zn-SOD in the baculovirus expressionthe proteins using a Zeeman atomic absorption system In order to solve this problem, we produced recombinant proteins and investigated enzymatic characteristics of the mutant proteins [33]. We chose a baculovirus expression system for overproduction and characterization of mutant SODs because, from an evolutionary point of view, insect cells are closer to mammalian cells than Escherichia coli or yeast. Sf21 insect cells infected with a baculovirus carrying the human Cu,Zn-SOD cdna produced a large amount of Cu,Zn-SOD apoprotein in the conventional medium. The SOD activity of the apoprotein, which was initially very low, increased when Cu 2+ was added to the culture medium in a dose-dependent manner. The protein produced by the infected cells was purified by a simple procedure involving two chromatographic steps: DE52 ion exchange and gel filtration. Identity of the recombinant Cu,Zn-SOD to the human enzyme was confirmed by immunochemical reactivity and by partial amino acid sequencing of peptides from the purified protein (50 amino acid residues in total). Enzymatic characterization of mutant Cu.Zn-SODs We made wild-type and six mutant Cu,Zn-SODs, G37R, G41D, G41S, H43R, G85R and I113T, and compared their activities. All the SODs showed slightly lower activities, 60-80% of the wild-type. The molecular sizes were the same except for the Gly85 to Arg mutation. Because Cu,Zn-SOD is very stable enzyme with resistance to high temperature, various chemicals, including metal-chelating reagents, and spectrophotometer after separation of proteins by filtration, and by colorimetry using bathocuproine. While atomic absorption detected quite a small amount of the copper compared with the inactivated enzymes, the amount of copper ions determined by the bathocuproine method correlated well with the amount of the inactivated SOD. This suggests that copper ions still bind weakly to the polypeptide after inactivation by hydrogen peroxide. Although several reports have suggested that the stability of mutant Cu,Zn-SODs is decreased [24], it is not clear which factor is actually involved in their destabilization. We therefore investigated the enzyme activity of the mutant SODs after incubation under various conditions. We first treated mutant as well as wild-type enzymes with 1 mm hydrogen peroxide, and measured SOD activity. With this treatment, mutant SODs were more severely inactivated than wild-type enzyme. Although Cu,Zn-SOD scavenges superoxide, the resultant hydrogen peroxide can be converted to a hydroxyl radical by either the Fenton-type reaction in the presence of free transition metal ions or interaction with Cu,Zn-SOD itself [31,32]. We examined the hydroxyl radical production in the presence of various concentrations of Cu 2+ and/or wild-type SOD by ESR. Production of a hydroxyl radical in the absence of SOD increased in a Cu 2+ concentration-dependent manner up to a maximum of 300 \xm. In the presence of wild-type SOD, increased concentrations of hydroxyl radicals were produced at all Cu 2+ concentrations, even under the condition without Cu 2+ in the presence of 5 mm EDTA. Thus, under the condition with free copper, which could be provided from inactivated

38 J. Fujii et al. [A] H202 Protein Cleavage Cu,Zn-SOD Glucose Lipid Peroxidation n [B] Mutant SOD G1^ose Inactivation Fig. 1. (A) Schematic diagram of the effects of Cu,Zn-SOD glycation on proteins, DNA and lipids through production of ROI in the cells. (B) A possible mechanism for inactivation of mutant Cu,Zn-SODs by glycation and the subsequent development of FALS. mutant SOD in FALS, more hydroxyl radical can be formed which can damage cells. We next examined the correlation between hydroxyl radical production in the presence of 1 mm hydrogen peroxide and 5 mm EDTA and SOD activity of mutant enzymes. The amount of hydroxyl radicals produced by mutant enzymes was the same as or less than that of wild-type enzyme, and hydroxyl radical formation and SOD activity were well correlated (r = 0.85). Thus, in terms of hydroxyl radical production, there is no gain of function in mutant SODs, which is suggested by the data from the study of transgenic mice [28,29]. The controversial data may be explained by the same mechanism as described in the cardiovascular system, where both higher and lower levels of SODs exert cytotoxicity in the oxygenated heart [34]. Effect of glycation on mutant Cu.Zn-SOD There are several reports indicating that the incidence of glucose intolerance is increased in ALS (~40% of patients) and the glycation reaction is advanced [35]. Since the glycation reaction is accelerated under high blood glucose conditions, it is possible that mutant SODs are also glycated and suffer from damage in FALS patients. Hence, we investigated the effect of glycation on the mutant Cu,Zn-SOD stability in vitro.

Glycation of Cu,Zn-SOD 39 Incubation with 0.1 M glucose for 7 or 14 days induced SOD inactivation more markedly in mutant enzymes than the wild-type enzyme, suggesting that glycation is a factor for the inactivation of mutant SODs in FALS. The Amadori products in mutant enzymes with substitution to Arg residues seemed to be greater according to the reactivity to anti-glucitol lysine antibody [36]. This would facilitate the inactivation of the mutant enzymes. G85R mutant was most severely inactivated by this incubation. Using G85R mutant SOD, we examined the effects ofvarying concentrations of glucose on the SOD activity. The inactivation occurred in a glucose concentration-dependent manner. Because inactivation occurred fairly quickly, glucose would affect the structure of the mutant proteins more severely than the wild-type enzyme (Figure 1). We also examined the effect of fructose and the hydroxyl radical scavenger sodium formate on the inactivation of SOD and confirmed that the inactivation was significantly reduced in the presence of sodium formate. These findings also suggest the involvement of a hydroxyl radical in the inactivation of the mutant SOD by the glycation reaction, as demonstrated. Acknowledgements. This work was supported in part by a Grant-in-Aid for Scientific Research on Priority Areas by the Ministry of Education, Science, and Culture, Japan and by the Mochida Memorial Foundation for Medical and Pharmaceutical Research. References 1. Monnier VM. Toward a Maillard reaction theory of aging. In: Baynes JW, Monnier VM, eds. The Maillard Reaction in Aging, Diabetes and Nutrition. Alan R. Liss, New York, 1989; 1-22 2. Arai K, Iizuka S, Tada Y, Oikawa K, Taniguchi N. Increase in the glucosylated form of erythrocyte Cu,Zn-superoxide dismutase in diabetes and close association of the nonenzymatic glucosylation with the enzyme activity. Biochim Biophys Ada 1987; 924: 292-296 3. Arai K, Maguchi S, Fujii S, lshibashi H, Oikawa K, Taniguchi N. Glycation and inactivation of human Cu-Zn-superoxide dismutase. Identification of the in vitro glycation sites. J Biol Chem 1987; 262: 16969-16972 4. Kondo T, Murakami K, Ohtuka Y, Tsuji M, Gasa S, Taniguchi N, Kawakami Y. Estimation and characterization of glycosylated carbonic anhydrase I in erythrocytes from patients with diabetes mellitus. Clin Chim Ada 1987; 166: 227-236 5. Shilton BH, Walton DJ. Sites of glycation of human and horse liver alcohol dehydrogenase in vivo. J Biol Chem 1991; 266: 5587-5592 6. Takahashi M, Lu Y, Myint T, Fujii J, Wada Y, Taniguchi N. In vivo glycation of aldehyde reductase, a major 3-deoxyglucosone reducing enzyme. Identification of glycation sites. Biochemistry 1995; 34: 1433-1438 7. Garner MH, Bahador A, Sachs G. Nonenzymatic glycation of Na.K-ATPase: effect of ATP hydrolysis and K+ occlusion. J Biol Chem 1990; 265: 15058-15066 8. Sakurai T, Tsuchiya S. Superoxide production from nonenzymatically glycated protein. FEBS Lett 1988; 236:406-410 9. Mullarkey CJ, Edelstein D, Bronlee M. Free radical generation by early glycation products: a mechanism for accelerated atherogenesis in diabetes. Biochem Biophys Res Commun 1990; 173: 932-939 10. Takahashi M, Fujii J, Teshima T, Suzuki K, Shiba T, Taniguchi N. Identity of a major 3-deoxyglucosone-reducing enzyme with aldehyde reductase in rat liver established by amino acid sequencing and cdna expression. Gene 1995; 127: 249-253 11. Tomlinson DR, Stevens EJ, Diemel CT. Aldose reductase inhibitors and their potential for the treatment of diabetic complications. Trends Pharmacol Sci 1994; 15: 293-297 12. Takahashi M, Fujii J, Miyoshi E, Hoshi A, Taniguchi N. Elevation of aldose reductase gene expression in rat primary hepatoma and hepatoma cell lines: implication in detoxification of cytotoxic aldehydes. Int J Cancer 1995; 62: 749-754 13. Takahashi M, Hoshi A, Fujii J et al. Induction of aldose reductase gene expression in LEC rats during the development of hereditary hepatitis and hepatoma. Jpn J Cancer Res in press 14. Taniguchi N. Clinical significances of superoxide dismutases: changes in aging, diabetes, ischemia, and cancer. Adv Clin Chem 1992; 29: 1-59 15. Kawamura N, Ookawara T, Suzuki K, Konishi K, Mino M, Taniguchi N. Increased glycated Cu,Zn-superoxide dismutase levels in erythrocytes of patients with insulin-dependent diabetes mellitus. J Clin Endocrinol Metab 1992; 74: 1352-1354 16. Ookawara T, Kawamura N, Kitagawa Y, Taniguchi N. Sitespecific and random fragmentation of Cu,Zn-superoxide dismutase by glycation reaction. Implication of reactive oxygen species. J Biol Chem 1992; 267: 18505-18510 17. Islam KN, Takahashi M, Higashyama S et al. Fragmentation of ceruloplasmin following non-enzymatic glycation reaction. J Biochem 1995; 118: 1054-1060 18. Kaneto H, Fujii J, Suzuki K. et al. DNA cleavage induced by glycation of Cu,Zn-superoxide dismutase. Biochem J 1994; 304: 219-225 19. Rosen, DR, Siddique T, Patterson D et al. Mutations in Cu/Zn-superoxide dismutase gene are associated with familial amyotrophic lateral sclerosis. Nature 1993; 362: 59-62 20. Nakano R, Sato S, Inuzuka T et al. A novel mutation in Cu/Zn superoxide dismutase gene in a Japanese familial amyotrophic lateral sclerosis. Biochem Biophys Res Commun 1994; 200: 695-703 21. Hirano M, Fujii J, Nagai Y el al. A new variant Cu/Zn superoxide dismutase (Val7 -» Glu) deduced from lymphocyte mrna sequences from Japanese patients with familial amyotrophic lateral sclerosis. Biochem Biophys Res Commun 1994; 204: 572-577 22. Brown RH, Jr. Amyotrophic lateral sclerosis: recent insights from genetics and transgenic mice. Cell 1995; 80: 687-692 23. Deng H-X, Hentati A, Tainer JA et al. Amyotrophic lateral sclerosis and structural defects in Cu.Zn superoxide dismutase. Science 1993; 261: 1047-1051 24. Borchelt DR, Lee MK, Slunt HS et al. Superoxide dismutase 1 with mutations linked to familial amyotrophic lateral sclerosis possesses significant activity. Proc Natl Acad Sci USA 1994; 91: 8292-8296 25. Rabizadeh S, Gralla EB, Borchelt DR et al. Mutations associated with amyotrophic lateral sclerosis convert superoxide dismutase from an antiapoptotic gene to proapoptotic gene: studies in yeast and neuronal cells. Proc Natl Acad Sci USA 1995; 92: 3024-3028 26. Rothstein JD, Bristol LA, Hosier B, Brown RH, Jr, Kuncl RW. Chronic inhibition of superoxide dismutase produces apoptotic death of spinal neurons. Proc Natl Acad Sci USA 1994; 91: 4155-4159 27. Troy CM, Shelanski ML. Down-regulation of copper/zinc superoxide dismutase causes apoptotic death in PCI 2 neuronal cells. Proc Natl Acad Sci USA 1994; 91: 6384-6387 28. Gurnery ME, Pu H, Chiu AY et al. Motor neuron degradation in mice that express a human Cu.Zn-superoxide dismutase mutation. Science 1994; 264: 1772-1775 29. Ripps ME, Huntley GW, Hof PR, Morrison JH, Gordon JW. Transgenic mice expressing an altered murine superoxide dismutase gene provide an animal model of amyotrophic lateral sclerosis. Proc Natl Acad Sci USA 1995; 92: 689-693 30. Yim MB, Chock PB, Stadtman ER. Copper,zinc superoxide dismutase catalyzes hydroxyl radical production from hydrogen peroxide. Proc Natl Acad Sci USA 1990; 87: 5006-5010 31. Yim MB, Chock PB, Stadtman ER. Enzyme function of copper.zinc superoxide dismutase as a free radical generator. J Biol Chem 1993; 268:4099-4105 32. Sato K, Akaike T, Kohno M, Ando M, Maeda H. Hydroxyl

40 J. Fujii et al. radical production by H2O2 plus Cu.Zn-superoxide dismutase Cu.Zn-superoxide dismutase is lost at high doses in the reflects the activity of free copper released from the oxidatively oxygenated heart. Free Rad Biol Med 1990; 4: 9-14 damaged enzyme. J Biol Chem 1992; 267: 25371-25377 35. Poulton KR, Rossi ML. Peripheral nerve protein glycation and 33. Fujii J, Myint T, Seo HG, Kayanoki Y, Ikeda Y, Taniguchi N. muscle fructolysis: evidence of abnormal carbohydrate metab- Characterization of wild-type and amyotrophic lateral sclerosis- olism in ALS. Fund Neurol 1993; 8: 33-42 related mutant Cu,Zn-superoxide dismutases overproduced in 36. Myint T, Hoshi S, Ookawara T, Miyazawa N, Suzuki K, baculovirus-infected insect cells. J Neurochem 1995; 64: Taniguchi N. Immunological detection of glycated proteins in 1456-1461 normal and streptozotocin-induced diabetic rats using anti 34. Omar BA, Gad NM, Jordan MC et al. Cardioprotection by hexitol-lysine IgG. Biochim Biophys Ada 1995; 1272: 73-79