Inhibition of glyceraldehyde-3-phosphate dehydrogenase by peptide and protein peroxides generated by singlet oxygen attack Philip E. Morgan 1 , Roger T. Dean 2 and Michael J. Davies 1 1 EPR and 2 Cell Biology Groups, The Heart Research Institute, Sydney, New South Wales, Australia Reaction of certain peptides and proteins with singlet oxygen (generated by visible light in the presence of rose bengal dye) yields long-lived peptide and protein peroxides. Incubation of these peroxides with glyceraldehyde-3-phosphate dehy- drogenase, in the absence of added metal ions, results in loss of enzymatic activity. Comparative studies with a range of peroxides have shown t hat this inhibition is concentration, peroxide, a nd time dependent, with H 2 O 2 less efficie nt than some peptide peroxides. Enzyme inhibition correlates with loss of both the peroxide and e nzyme thiol residues, with a stoichiometry of two thiols lost per peroxide c onsumed. Blocking the thiol residues prevents reaction with the per- oxide. This stoichiometry, the lack of metal-ion dependence, and the absence of electron paramagnetic resonance (EPR)- detectable species, is consistent with a molecular (nonradi- cal) reaction between the active-site thiol of t he enzyme and the peroxide. A number of low-molecular-mass compounds including thiols and ascorbate, but not Trolox C, can pre- vent inh ibition by removing the initial peroxide, or s pecies derived from it. In contrast, g lutathione reductase and lac- tate dehydrogenase are poorly inhibited by these peroxides in the absence of added Fe 2+ –EDTA. The presence of this metal-ion complex enhanced the inhibition observed with these enzymes consistent with the o ccurrence of radical- mediated reactions. Overall, these studies demonstrate that singlet oxygen-mediated damage to a n i nitial target protein can r esult in selective subsequent damage to other proteins, as evidenced by loss of enzymatic activity, via the formation and subsequent reactions of protein peroxides. These reac- tions may be important in the development o f cellular d ys- function as a result of photo-oxidation. Keywords: protein oxidation; protein peroxides; protein radicals; s inglet oxygen; photo-oxidation. Singlet oxygen (molecular o xyge n in its 1 D g state; 1 O 2 )is generated by a number of enzymatic and chemical reactions, by UV exposure, and by visible light in the presence of a number of exogenous or endogenous cellular sensitisers. 1 O 2 generation has been reported in myeloperoxidase- and eosinophil peroxidase-catalysed reactions [1–3], and by some activated cell types including neutrophils [4], eosino- phils [3,5], and macrophages [6]. As a result of the wide- spread exposure of humans to UV and visible light, 1 O 2 has been sugge sted t o p lay a key role in the development of a number of human pathologies including cataract, sunburn, some skin cancers and aging [7–12]. 1 O 2 reacts with a range of biological molecules including DNA [13,14], cholesterol [15,16], lipids [15,17,18], and amino acids and proteins [12,19,20]. Proteins are major biological targets as a result of their abundance and high rate constants for reaction [21], with damage occurring primarily at Trp, Met, Cys, His and Tyr side-chains [12,19,20]. Reaction with Trp, H is and T yr residues h as been shown t o yield peroxides, although the structure of some of these materials remain s to be fully established (reviewed in [12,19,20]). Previous studies have identified the C-3 site on the indole ring of Trp as a major site of peroxide formation [22], and our recent studies have demonstrated that the major peroxide generated with Tyr residues is a ring-derived, C-1, dieneone hydroperoxide (A. Wright, W. A. Bubb, C. L. Hawkins & M. J. Davies, unpublished results). Further species are also formed with free Tyr [23]. Both endo- and hydro-peroxides have been reported with His [24]. 1 O 2 -mediated oxidation of proteins also yields peroxides, with Tyr, Trp and His residues likely targets [25]. All of these peroxides are unstable in solution, with decomposition enhanced by reducing agents, UV light and metal ions ([25]; A. W right, W. A. Bubb, C. L. Hawkins & M. J. Davies, unpublished results). Reaction with some metal ions generates radical species ([25]; A. Wright, C. L. Hawkins & M. J. Davies, unpublished r esults). Previous studies with protein peroxides generated by high-energy radiation (e.g. c-sources, X-rays), meta l ion/ peroxide systems, thermal sources of peroxyl radicals, peroxynitrite, and activated white cells [26,27], have shown that these species play a key role in the propagation of oxidative chain reactions within proteins [12,28]. These species can oxidize other biomolecules, including lipids, Correspondence to M. J. Davies, EPR Group, The Heart Research Institute, 145 Missenden Road, Camperdown, Sydney, New South Wales 2050, Australia. Fax: + 61 29550 3302, E-mail: m.davies@hri.org.au Abbreviations: EPR, electron paramagnetic resonance; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GR, glutathione reduc- tase; GSH, reduced glutathione; LDH, lactate dehydrogenase; 2MPG, N-(2-mercaptopropionyl)glycine; N-Ac-Trp-OMe, N-acetyl trypto- phan methyl ester; N-Ac-Trp-OMe-OOH, peroxides formed on N-acetyl tryptophan methyl ester by reaction with 1 O 2 ;NEM, N-ethylmaleimide; 1 O 2 , molecular oxygen in its first excited singlet ( 1 D g ) state; PBN, N-t-butyl-a-phenylnitrone. Enzymes: glyceraldehyde-3-phosphate dehydrogenase (EC 1.2.1.12); glutathione reductase (EC 1.6.4.2); lactate dehydrogenase (EC 1.1.1.27) . Note: a website is available at www.hri.org.au (Received 13 N ovember 2001, revised 12 February 2002, accepted 20 February 2002) Eur. J. Biochem. 269, 1916–1925 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.02845.x antioxidants and DNA [27,29–31], with some of these reactions involving peroxide-derived radicals [30–33]. The reactions of protein peroxides with other proteins have not been investigated in depth, and would be expected to be distinct from the reactions of low-molecular-mass alkyl a nd lipid peroxides. A preliminary report has appeared on the inhibition of glutathione reductase by radiation-generated protein peroxides [34], and evidence presented for inhibition of enzymes, by radiation-generated species, in erythrocytes [35]. Such results cannot be extrapolated to 1 O 2 -generated species, owing to their different sites and chemistries ([25– 27]; A. Wright, C. L. Hawkins & M. J. Davies, unpublished results). In this study we have examined the inhibition of two thiol- dependent enzymes [glyceraldehyde-3-phosphate dehydro- genase (GAPDH) and glutathione reductase (GR)], and one nonthiol-dependent enzyme (lactate dehydrogenase, LDH), by 1 O 2 -generated peroxides f ormed on p eptides and proteins. These data have been compared with those obtained u sing H 2 O 2 . The role of radical vs. nonradical processes has been investigated, as has the prevention of such damage. MATERIALS AND METHODS Amino acids, peptides and antioxidants were commercial samples of high purity. BSA (fraction V, > 98%), lysozyme (chicken egg white, % 95 %), RNase A (bovine pancreas, essentially protease and salt free), GR [bakers yeast, in 3.6 M (NH 4 ) 2 SO 4 , pH 7 .0], GAPDH (rabbit muscle, lyophilized powder) and LDH [rabbit muscle, in 3.2 M (NH 4 ) 2 SO 4 , pH 6.0, or lyophilized powder] were from Sigma. GAPDH [rabbit muscle, in 3.2 M (NH 4 ) 2 SO 4 +0.1m M EDTA, pH 7.5], NADPH (tetrasodium salt), NAD + (free acid), NADH (disodium salt) were f rom B oehringer Mannheim. These preparations did not contain materials that interfered with peroxide formation or e nzyme activity measurements. The water used was passed through a four-stage Milli Q system equipped with a 0.2-lm-pore-size final filter. Solu- tions of Fe 2+ –EDTA (1 : 1 complex) were prepared using de-oxygenated water and maintained under oxygen-free N 2 . All concentrations given are final values. Peroxides were generated on BSA, lysozyme, RNase A (each 50 mgÆmL )1 ) and the peptides N-acetyl tryptophan methyl ester (N-Ac-Trp-OMe), Gly-His-Gly and Gly-Tyr- Gly (each 2.5 m M ), by photolysis with visible light (from a Kodak S-AV 2050 slide projector) through a 345-nm cut-off filter in the presence of 10 l M rose bengal dye ([25]; A. Wright, C. L . Hawkins & M. J. Davies, unpublished results). Solutions were kept on ice during photolysis (30 min BSA, 60 min for lysozyme and peptides, 120 min for RNase A), and were continually aerated. After cessation of photolysis, c atalase (Sigma, bovine liver, 5 lgÆmL )1 for BSA, 50 lgÆmL )1 for lysozyme and RNase A, 250 lgÆmL )1 for p eptides) was added, unless stated otherwise, to remove H 2 O 2 and the samples incubated for 30 min at room temperature before freezing ()80 °C) in aliquots. Peroxide concentrations were determined by a modified FOX (FeSO 4 /xylenol orange) assay, using H 2 O 2 standards [36]. This assay gives s imilar values to iodometric analysis (A. Wright, C. L. Hawkins & M. J. Davies, unpublished results). The effects of reductants on peroxides were determined by incubation of the samples with an approx- imately 20-fold excess of NaBH 4 over protein concentration for 30 m in at room temperature. Control samples were incubated under identical conditions. I mmediately post- incubation, samples were separated on a PD-10 column (Pharmacia), the protein fractions collected, and residual peroxides determined after correction for sample dilution (assessed by A 280 values). Thiol groups on GAPDH (2 mgÆmL )1 ) were blo cked by incubation for 30 min at room temperature with a 10-fold excess (over protein thiol concentration) of N-ethylmalei- mide (NEM) followed by separation of the treated protein from excess reagent by PD-10 chromatography, with water elution. Control samples were incubated in the absence of NEM. Protein concentration after PD-10 chromatography was determined by the BCA assay, using BSA standards (Pierce). Free thiol concentrations were assessed by incubation of 0.1 m gÆmL )1 GAPDH with 500 l M 5,5¢-dithiobis(2-nitro- benzoic acid) (in 100 m M phosphate buffer, pH 7.4) for 30 min at room temperature, with quantification of the released 5-thionitrobenzoic acid (TNB) anion measured using its absorbance at 412 nm and e 13 600 M )1 Æcm )1 [37]. Electron paramagnetic resonance (EPR) samples were prepared by addition of peroxide (200 l M ) to the enzyme in the presence of the spin trap N-t-butyl-a-phenylnitrone (PBN) (9.4 m M in 50 m M phosphate buffer, pH 7.4). Fe 2+ – EDTA (100 l M , 1 : 1 complex) was added where stated. Samples were incubated for 5 min at 20 °Cbeforeexami- nation in a standard, fl attened, aqueous solution cell (WG- 813-SQ; Wilmad, B uena, NJ, USA) using a Bruker EMX X-band spectrometer equipped with 100 kHz modulation and a cylindrical ER4103TM cavity. Typical spectrometer settings were: gain 1.0 · 10 6 , modulation amplitude 0.1 mT, time constant 163.8 ms, sweep time 81.9 s, centre field 348.0 mT, field sweep width 8.0 mT, microwave pow er 25.0 m W, frequency 9 .7 GHz, with four acquisitions aver- aged. GR ac tivity was examined at 37 °Cin50m M phosphate buffer, pH 7.4, containing 0.025 UÆmL )1 GR, 20 l M Fe 2+ – EDTA (where indicated), peroxide (20–200 l M ), and antioxidant (10-fold excess over peroxide concentration). Aliquots were removed as indicated, and residual activity assayed by the sequential addition of oxidized glutath ione (1 m M ) and NADPH (0.1 m M ), with consumption of the latter measured at 340 nm and 37 °C over t he period from 1.2 to 3.0 min after the addition of NADPH. LDH activity was assessed in a similar manner, except using 0.05 UÆmL )1 enzyme, with pyruvic acid (1 m M )andNADH(0.1m M ) added to the aliquots, and the loss of the latter monitored at 340 nm. GAPDH activity was assessed in 50 m M pyro- phosphate buffer, pH 7.4, with 0.15 U mL )1 enzyme and 15–100 l M peroxides. Aliquots were removed as indicated, glyceraldehyde-3-phosphate (1 m M ), NAD + (0.5 m M ), and sodium arsenate (25 m M , in water) added, and NADH formation monitored at 340 nm. All experiments were performed in duplicate or greater. Enzyme inhibition induced by the peptide and protein peroxides was compared to controls containing the c orres- ponding nonoxidized substrate. Statistical analyses com- paring multiple enzyme activities were performed using a one-way ANOVA and Dunnett’s posthoc test. Other analyses comparing multiple conditions were performed using a one-way ANOVA and Newman–Keuls posthoc test. Where Ó FEBS 2002 Enzyme inhibition by 1 O 2 -mediated protein peroxides (Eur. J. Biochem. 269) 1917 only one condition was compared to its corresponding control a Student’s pooled two-sample t-test was used. In all cases significance was assumed if P < 0.05. RESULTS Formation of peptide and protein peroxides Photolysis of solutions containing rose bengal and N-Ac- Trp-OMe, Gly-His-Gly, Gly-Tyr-Gly, lysozyme or RNase A with visible light (k > 345 nm) in the presence of O 2 resulted in the generation of peroxides as detected by a modified FOX assay [25,36]. Treatment of such samples with catalase, after the cessation of illumination, resulted, in most cases, in a very rapid initial decrease in peroxide concentration, and a subsequent slow decay (Fig. 1). The absolute levels of peroxide lost in the rapid phase after addition of catalase, was substrate dependent (Fig. 1). The fast initial loss is ascribed to the removal of H 2 O 2 generated during the photolysis (e.g [38]). The subsequent slow decay is a ssigned to thermal decomposition of peptide or protein peroxides ([25]; reviewed in [12,19,22]). In the case of lysozyme (Fig. 1C) only thermal decompo- sition of protein peroxides is evident, as no H 2 O 2 appears to be formed during the photolysis. The presence of peroxide groups on the proteins tested was confirmed by the coelution of the FOX assay-positive material with the protein containing fractions from size-exclusion chroma- tography columns (data not shown). High concentrations of catalase (£ 250 lgÆmL )1 ), and a 30-min preincubation period, were employed in all subsequent experiments to ensure complete, and rapid, removal of H 2 O 2 before the peroxides were u sed in other experiments. Omission of the rose bengal, photolysis in the absence of O 2 , or incubation of complete samples in the absence of light, resulted in peroxide concentrations of < 5 l M (data not shown). Stability of peptide and protein peroxides The decay of the peroxides generated on N-Ac-Trp-OMe, Gly-Tyr-Gly, Gly-His-Gly, lysozyme and RNase A was studied over time at 37 °C (Fig. 2). Addition of GAPDH, to these incubations resulted in a significantly (P < 0.01) enhanced rate of decay of the peroxides formed o n N-Ac- Trp-OMe, Gly-Tyr-Gly, and Gly-His-Gly consistent with reaction of these peroxides with this enzyme (Fig. 2 ). Similar experiments with lysozyme - and RNase A -derived peroxides did not yield statistically significant data (data not shown). Experiments with LDH i n the place o f GAPDH, and the same peroxides, did not result in a statistically enhanced rate of decay of N-Ac-Trp-OMe and Gly-Tyr-Gly peroxides (P > 0.05). An enhanced rate of decomposition was detected with Gly-His-Gly peroxide (%30% after 30 min compared to 12% in controls, P < 0.01), though this was much less marked than that observed with GAPDH (data not shown). Analogous experiments were not carried out with GR due to the quantity of material required. Treatment of and RNase A peroxides with NaBH 4 (approx. 20-fold molar excess relative to protein concentra- tion) for 3 0 min at room temperature, and subsequent separation of the treated protein from excess reductant by PD-10 chromatography, resulted in the loss of > 97% of the initial peroxides. This is in accord with previous studies [25,26]. Treatment with Fe 2+ –EDTA, reduced glutathione (GSH), ebselen and other thiols a lso rapidly removes such peroxides (data not shown; A. Wright, C. L. Hawkins & M. J. Davies, unpublished results; P. E. Morgan, R. T. Dean & M. J. Davies, unpublished data). Similar studies were not carried out with peptide-derived peroxides as excess reductant, w hich interferes with the peroxide assay, Fig. 1. Formation of H 2 O 2 and peptide and protein peroxides after photo-oxidation with 1 O 2 generated by rose bengal i n the presence of visible light and oxygen. (a) RNase A (50 mgÆmL )1 ); (b) N-Ac-Trp- OMe (2.5 m M ); and (c) lysozyme (50 mgÆmL )1 )werephotolysedwith visible light for 60 min in the presence of rose b engal (10 l M )with continuous gassing with air at 4 °C. Immediately after the cessation of photolysis catalase w as added to part of the sample (50 lgÆmL )1 for proteins, 250 lgÆmL )1 for N-Ac-Trp-OMe) and peroxide levels assayed at the indicated times using a modified FOX assay. (¤) Catalase added; (h) no catalase added. Initial peroxide concentrations were in the range of 420–520 l M for RNase A, 540–660 l M for N-Ac-Trp-OMe, and 130–180 l M for lysozyme. Data are means ± S D; where no error bar is visible it is obscured by the symbol. Statistical analysis between conditions was by a Student’s pooled two- sample t-test, ** P < 0.01. 1918 P. E. Morgan et al. (Eur. J. Biochem. 269) Ó FEBS 2002 could not be readily removed from the reaction mixture, and lysozyme was unable to be analysed as the protein precipitated on addition of NaBH 4 . Interaction of GAPDH thiols with peptide peroxides The interaction of free thiol groups on GAPDH with N-Ac- Trp-OMe peroxides was assessed by measurement of the change in GAPDH thiol concentration on incubation with this peroxide. Figure 3A shows that as the peroxide concentration decreased, a concomitant, time-dependent, decrease in thiol concentration was observed . The concen- tration of thiols lost (24.9 ± 1.2 l M at 30 min), is approxi- mately double that of the peroxides lost under identical conditions (11.2 ± 1.1 l M at 30 min, Fig. 2A), consistent with a stoichiometry of two thiol groups lost per peroxide molecule consumed. Further evidence for an interaction of the thiol groups on GAPDH with N-Ac-Trp-OMe peroxides was obtained by pretreatment of the GAPDH with NEM. This resulted in Fig. 2. Thermal decay of peptide peroxides over time at 37 °Cinthe absence or presence of added GAPDH. Peptide peroxide samples were generated as described in Fig. 1 a nd the text. Im mediately after c es- sation of photolysis catalase was added and the samples incubated for 30 min at room temperature. The residual peroxide levels after further incubation at 37 °C, were m easured at the indicated times for either untreated controls (d), or samples with added GAPDH (n; 1mgÆmL )1 ), using a modified FOX assay. (A) N-Ac-Trp-OMe per- oxides; ( B) Gly-His-Gly peroxides; (C) Gly-Tyr-Gly peroxides. In all cases the initial (postcatalase treatment) peroxide concentration in the incubation mixtures was 20 l M . Data are means ± SD; where no error bar is visible it is obscured by the symbol. Statistical analysis was by a Student’s pooled two-sample t-test, ** P <0.01. Fig. 3. (A) Loss of thiol groups present o n GAPDH on incubation with N-Ac-Trp-OMe peroxides, and (B) Effect of blocking the free thiols groups on GAPDH on loss of N-Ac-Trp-OMe peroxides. (A) GAPDH (1 m gÆmL )1 ) was incubated with 20 l M N-Ac-Trp-OMe peroxides for 30 min at 37 °C, with the concentration of free GAPDH thiols mea- sured by reaction with 5,5¢-dithiobis(2-nit robe nzoic a cid) at the indi- cated times. (r) GAPDH in presence of N-Ac-Trp-OMe peroxides; (h) GAPDH in the absence of added peroxide. Initially, 7 .3 ± 0.2 free thiols p er GAPDH tetramer were detected. Data are means ± SD; where no e rror bar is visible it is obscured by the symbol. Statistical analysis was by a Student’s pooled two-sample t-test, ** P < 0.01. (B) GAPDH was incubated for 30 min at room temp with, or without NEM, followed by PD-10 column treatment to re-isolate the enzyme and remove excess reagent (c ontrol, 42 ± 6% of thiols free; N EM- treated, 20 ± 3% of thiols free). The re-isolated enzyme was then incubated with 20 l M N-Ac-Trp-OMe peroxides at 37 °C, and the residual peroxide levels measured at the indicated times using a modi- fied FOX assay. (j) Perox ide loss observed in prese nce of control (non-NEM treated) GAPDH ; (·) peroxide loss observed in presence of NEM-treated GAPDH; ( d) peroxide l oss in absence of adde d GAP- DH (cf. Fig. 2A). Data are means ± S D; where no error bar is visible it is obscured by the symbol. Statistical analysis was by one-way ANOVA with Newman–Keuls posthoc test; unlike letters indicate statistically distinct results at the P <0.05level. Ó FEBS 2002 Enzyme inhibition by 1 O 2 -mediated protein peroxides (Eur. J. Biochem. 269) 1919 the blocking of % 50% of the free thiols on the enzyme when compared to controls. Complete blocking of all thiol groups was not attempted a s the requirement for high concentra- tions of NEM can result in other modifications [39]. Subsequent incubation of such NEM-treated GAPDH with N-Ac-Trp-OMe peroxides resulted in a much slower, and less dramatic, loss in peroxide concentration compared to the non-NEM treated control (Fig. 3B), confirming that the thiol groups on GAPDH play a role in the peroxide loss. Similar experiments were not carried out with other peroxides or enzymes. Enzyme inhibition studies Peroxides g enerated on N-Ac-Trp-OMe, Gly-Tyr-Gly, Gly-His-Gly, lysozyme and RNase A were incubated with GAPDH, GR, and LDH, in the presence and absence of added Fe 2+ –EDTA and the residual enzymatic activity determined (Fig. 4, Table 1). Lower concentrations of these enzymes were employed in these studies, c ompared to those reported above, to prevent substrate depletion. Comparative studies were also carried out with H 2 O 2 . The EDTA complex of Fe 2+ was employed to prevent potential binding of Fe 2+ to the target enzymes; omission of the EDTA resulted in less efficient enzyme inhib ition (data not shown). GAPDH w as rapidly inactivated, in a time-dependent manner, by all t he peroxides tested, in both the absence and presence of added Fe 2+ –EDTA. The rate of i nhibition of GAPDH by N-Ac-Trp-OMe peroxides was concentration dependent over the range tested (20–200 l M peroxide). Incubation of GAPDH with nonphotolysed samples of the peptide or proteins (with, o r without, 20 l M Fe 2+ –EDTA), or with Fe 2+ –EDTA alone, resulted in slow loss of enzyme activity, presumably owing to slow denaturation (Fig. 4A). GAPDH was readily inhibited by H 2 O 2 ,whichwas employed as a positive control, in either the presence, or absence, of Fe 2+ –EDTA. Comparison of the data obtained with H 2 O 2 and N-Ac-Trp-OMe peroxides showed that fivefold higher concentrations of H 2 O 2 needed to be employed to generate a similar r ate and extent of inhibition (Fig. 4 B). Inhibition by protein-derived peroxides was slower than that induced by the peptide peroxides at Fig. 4. Inhibition of glyceraldehyde-3-phosphate dehydrogenase on incubation with H 2 O 2 , and peptide- and protein-peroxides, in the presence and absence of added Fe 2+ –EDTA. GAPDH was incubated at 37 °C for 30 min with (a) N-Ac-T rp-OMe p eroxides (20 l M ); (b) H 2 O 2 (100 l M ); (c) Gly- His-Gly peroxides (15 l M ); (d) Gly-Tyr-Gly peroxides (20 l M ); (e) RNase A peroxides (100 l M ); and 120 m in with (f) lysozyme peroxides (20 l M ). Fe 2+ –EDTA (20 l M ) was added w here indicated. Control samples co ntained equal concentrations of nonphotolysed materials, or wate r (in the case of H 2 O 2 ). Activity is expressed as a percentage of that of the nonphotolysed (nonperoxide containing) samples without added Fe 2+ –EDTA. For further details see the Materials and methods. (·) Peroxide-containing samples in presence of added Fe 2+ –EDTA; (h) Peroxide-containing samples in absence of added Fe 2+ –EDTA; (n) nonp hotolysed/non H 2 O 2 containing sample s in presence of added Fe 2+ –EDTA; (r) n onpho- tolysed/non H 2 O 2 containing samples in ab sence of added Fe 2+ –EDTA. Statistical a nalyses (one-way ANOVA with Dunnett’s posthoc test) compared all conditions t o the nonphoto lysed/non -H 2 O 2 control without added Fe 2+ –EDTA, ** P < 0.01. Where n o error bar is visible it is obscured by the symbol. 1920 P. E. Morgan et al. (Eur. J. Biochem. 269) Ó FEBS 2002 identical peroxide concentrations demonstrating that per- oxide size and/or electronic charge p lay a role in d etermin- ing the rate of inhibition. Identical studies using GAPDH with samples of RNase A peroxides which had been pretreated with NaBH 4 to remove peroxides (see above), gave similar extents of inhibition to control, nonperoxide-containing RNase sam- ples, confirming the requirement for peroxide groups for enzyme inhibition. GR was inhibited on incubation with N-Ac-Trp-OM e peroxides, but only at the highest concentrations tested (200 l M ) (Table 1). This inhibition was not stimulated by added Fe 2+ –EDTA. In contrast, rapid inhibition of GR by H 2 O 2 was only observed in the presence of Fe 2+ –EDTA (Table 1). LDH was not inhibited by the highest concen- trations of peptide and protein peroxides (200 l M )tested,in either the presence or absence of Fe 2+ –EDTA (data not shown), whereas this enzyme was readily inhibited by H 2 O 2 in the presence, but not absence, of Fe 2+ –EDTA (Table 1). As with GAPDH, a slow loss of enzyme activity was observed, with both GR and LDH, in control samples; this has been ascribed to slow thermal i nactivation. Examination of the role of peroxide-derived radicals in enzyme inhibition To examine whether radical species were generated during the inactivation of GAPDH and L DH by peroxides in the absence of added metal ions, GAPDH (24 mgÆmL )1 )and LDH (6 m gÆmL )1 ) were incubated with the spin trap PBN (9.4 m M )andN-Ac-Trp-OMe peroxides or H 2 O 2 (both 200 l M ) for extended periods and examined by EPR spectroscopy. No radical adducts were detected above those detected in controls. Experiments were not carried out with GR owing to the quantity of material required. Previous studies have demonstrated the formation of radicals from these peptide and protein peroxides in the presence of Fe 2+ – EDTA ([25]; A. Wright, C. L. Hawkins & M. J. Davies, unpublished results). Prevention of enzyme inhibition induced by peptide peroxides A number of compounds protected GAPDH or GR against inactivation when these materials were coincubated with the enzymes and N-Ac-Trp-OMe p eroxides or H 2 O 2 (Table 2). GSH, N-(2-mercaptopropionyl)glycine, ascorbic acid and dithiothreitol (all 200 l M ) all offered highly significant protection against the inhibition of GAPDH induced by 20 l M N-Ac-Trp-OMe peroxides in the presence of 20 l M Fe 2+ –EDTA (Table 2, cf. Figure 4A). Methionine, at an identical concentration, had a much less marked, although still statistically significant, effect. Trolox C was i neffective. All the compounds tested showed a significant protective effect at 2 m M in the LDH/Fe 2+ –EDTA/H 2 O 2 system (Table 2). In some cases, inclusion of these compounds in control samples resulted in minor changes in enzyme activity. Thus Trolox C caused a significant decrease in GR activity (P < 0.05), whilst blank experiments with added N-(2-mercaptopropionyl)glycine ( 2MPG) resulted in a significant increase in GR activity compared to the absence of this compound (P < 0.05). The latter effect is attributed to re-activation of inactive enzyme present in the sample. DISCUSSION Exposure of amino acids, peptides and proteins to radiation (ionizing, UV, or visible light in the presence of a photosensitiser) in the presence of O 2 , gives rise to peroxides [25–27,38]. With 1 O 2 , peroxides are formed primarily on Tyr, Trp and His side-chains [12,19,24,25,40]. Peroxides formed on N-Ac-Trp-OMe are primarily located at the C-3 site on the indole ring, and those on Gly-His-Gly and Gly- Tyr-Gly at ring positions on the His and Tyr side-chains, respectively ([19,23,24,40]; A. Wright, C. L. Hawkins & M. J. Davies, unpublished results). The location of such peroxides on proteins has yet to be fully determined. Recent s tudies have shown that protein peroxides a re also Table 1. Inhibition of glutathione reductase by H 2 O 2 and N-Ac-Trp-OMe peroxides, and lactate dehydrogenase by H 2 O 2 , in the presence and absence of added Fe 2+ –EDTA. Samples c ontaining glutathione reductase (0.025 UÆmL )1 ) were incubated at 37 °Cfor120minwithN-Ac-Trp-OMe peroxides (200 l M )andH 2 O 2 (200 l M ). Samples containing lactate dehydrogenase (0.05 UÆmL )1 ) were incubated at 37 °Cfor30minwithH 2 O 2 (200 l M ). 20 l M Fe 2+ –EDTA was present where indicated. Control solutions contain ed equal concentration s of nonphotolysed N-Ac-Trp-OMe, or water in the case of H 2 O 2 . Activity is expressed as a percentage of that of the nonphotolysed/non H 2 O 2 containing samples without added Fe 2+ – EDTA. Statistical analyses (one-way ANOVA with Dunnett’s p osthoc test ) compared all conditions to the nonphotolysed N-Ac-Trp-OMe/H 2 O control w ithout added Fe 2+ –EDTA; * P < 0.01. Enzyme Added agents Initial control activity (%) GR N-Ac-Trp-OMe (control) 87 ± 1 N-Ac-Trp-OMe + Fe 2+ –EDTA 84 ± 1 N-Ac-Trp-OMe + peroxides 68 ± 1* N-Ac-Trp-OMe + peroxides + Fe 2+ –EDTA 62 ± 4* GR H 2 O (control) 99 ± 1 H 2 O+Fe 2+ –EDTA 97 ± 4 H 2 O 2 98 ± 2 H 2 O 2 +Fe 2+ –EDTA 23 ± 5* LDH H 2 O (control) 72 ± 7 H 2 O+Fe 2+ –EDTA 70 ± 6 H 2 O 2 69 ± 2 H 2 O 2 +Fe 2+ –EDTA 10 ± 1* Ó FEBS 2002 Enzyme inhibition by 1 O 2 -mediated protein peroxides (Eur. J. Biochem. 269) 1921 generated in cells on exposure to 1 O 2 (A. Wright, C. L. Hawkins & M. J. Davies, unpublished results) or peroxyl radicals [41]. This study has shown, for the first time, that low concentrations of 1 O 2 -generated peptide- and protein-per- oxides can transmit damage from the initial site of oxidation to other cellular targets, and hence bring about chain oxidation reactions where oxidative damage to on e protein can r esult i n damage to multiple targets. It has been shown that these protein peroxides can inhibit GAPDH, which is a key cellular glycolytic enzyme, and GR, which r ecycles oxidized glutathion e and thereby maintains reducing equi- valents within the cell. Other enzymes, such as LDH, are unaffected, so such damage is selective. The long lifetime of these protein peroxides may allow these species to diffuse considerable distances from their site of formation, and hence induce damage at remote sites. The concentrations of peptide and protein peroxides that induce inhibition of GAPDH are similar to those which we have recently detected (% 20 l M ) on proteins in viable rose-bengal loaded THP-1 cells exposed to visible light (A. Wright, C. L. Hawkins & M. J. Davies, unpublished r esults). Previous studies (reviewed in [42]) have shown that GAPDH is rapidly, and specifically, inhibited in m yocytes, aortic endothelial and U937 (pro-monocyte) cells on exposure to H 2 O 2 in the absence of added metal ions [43– 45]. This inactivation arises via direct reaction of H 2 O 2 with a particularly reactive Cys residue (Cys149), which has a pK a of 5.4 owing to interaction with His176, in the a ctive site o f the e nzyme. This process gives a sulfenic acid (R-S- OH), which can be repaired by dithiothreitol. The isolated enzyme can also be inhibited by UV light [46], n itric oxide [37], superoxide radicals [42], ozone [47] and tert-butyl hydroperoxide [48]. Inhibition can also arise via radical Table 2. Percentage of enzyme activity retained after incubation of GAPDH, GR and LDH with N-Ac-Trp-OMe peroxides or H 2 O 2 at 37 °Cinthe absence, or presence, of a 10-fold excess (over peroxide concentration) of putative antioxidant. Samples containing GAPDH (0.15 UÆmL )1 )orLDH (0.05 UÆmL )1 ) were incubated for 30 min; those containing GR (0.025 UÆmL )1 ) for 120 min. A ll incubations contained 20 l M Fe 2+ –EDTA. Control samples contained nonphoto lysed N-Ac-Trp-OMe or H 2 O as appropriate. Statistical analyses (one-way ANOVA with Dunnett’s posthoc test) compared all conditions to the nonperoxide containing controls; * P <0.01,**P < 0.05. Enzyme Added agents Initial control activity (%) GAPDH + Fe 2+ –EDTA N-Ac-Trp-OMe (control) 86 ± 2 N-Ac-Trp-OMe peroxides 14 ± 2* N-Ac-Trp-OMe peroxides + GSH 83 ± 1 N-Ac-Trp-OMe peroxides + 2MPG 80 ± 1 N-Ac-Trp-OMe peroxides + Methionine 25 ± 4* N-Ac-Trp-OMe peroxides + Trolox C 13 ± 1* N-Ac-Trp-OMe peroxides + Dithiothreitol 87 ± 1 N-Ac-Trp-OMe peroxides + Ascorbic Acid 73 ± 3* GAPDH + Fe 2+ –EDTA H 2 O (control) 67 ± 5 H 2 O 2 7±3* H 2 O 2 + GSH 59 ± 2 H 2 O 2 + 2MPG 69 ± 5 H 2 O 2 + Methionine 2 ± 1* H 2 O 2 + Trolox C 4 ± 1* H 2 O 2 + Dithiothreitol 100 ± 3* H 2 O 2 + Ascorbic Acid 64 ± 4 GR + Fe 2+ –EDTA N-Ac-Trp-OMe (control) 88 ± 3 N-Ac-Trp-OMe peroxides 45 ± 1* N-Ac-Trp-OMe peroxides + GSH 74 ± 4** N-Ac-Trp-OMe peroxides + 2MPG 93 ± 2 N-Ac-Trp-OMe peroxides + Methionine 52 ± 2* N-Ac-Trp-OMe peroxides + Trolox C 44 ± 10* N-Ac-Trp-OMe peroxides + Dithiothreitol 93 ± 3 GR + Fe 2+ –EDTA H 2 O (control) 97 ± 4 H 2 O 2 23 ± 15* H 2 O 2 + GSH 88 ± 5 H 2 O 2 + 2MPG 109 ± 7 H 2 O 2 + Methionine 109 ± 6 H 2 O 2 + Trolox C 80 ± 1 H 2 O 2 + Dithiothreitol 55 ± 1* LDH + Fe 2+ –EDTA H 2 O (control) 70 ± 6 H 2 O 2 10 ± 1* H 2 O 2 + GSH 99 ± 8 H 2 O 2 + 2MPG 102 ± 10** H 2 O 2 + Methionine 95 ± 18 H 2 O 2 + Trolox C 112 ± 9* H 2 O 2 + Dithiothreitol 91 ± 5 H 2 O 2 + Ascorbic Acid 70 ± 6 1922 P. E. Morgan et al. (Eur. J. Biochem. 269) Ó FEBS 2002 reactions (e.g. involving HO Æ [46]. or O 2 – Æ [42,49]) that involve oxidation of Cys-149 to cysteic acid [47]. GR also contains an active site Cys r esidue [50]. GR is less-readily inhibited than GAPDH by H 2 O 2 , and loss of activity has been reported to require metal ions, be radical- mediated, involve oxidation of other residues in addition to the active site Cys (e.g. His467, Tyr114 and Trp residues [50]), and result in the formation of carbonyl groups [50–52]. A preliminary report has appeared on the inhibition of GR by radiation-generated peroxides [34]. LDH has been shown to be inhibited by a number of oxidants, with this requiring the p resence of metal ions [53], but is less sensitive to inhibition than GAPDH [42]. A similar pattern appears to hold with t he peptide and protein peroxides investigated in the current study, with GAPDH being more sensitive than GR and LDH, and inactivation of G APDH being metal- ion independent, whereas inhibition of GR and LDH by H 2 O 2 occurs most rapidly in the presence of metal ions. It is proposed that the inhibition of GAPDH and GR by these 1 O 2 -generated peptide and protein peroxides occurs via direct (nonradical) oxidation of the active site Cys residues (i.e. reaction 1, see below). This is s upported by the observations, with GAPDH, that thiol group loss occurs with similar kinetics to peroxide loss, and that blocking of % 50% of the thiol groups on the enzyme inhibits peroxide loss. Furthermore removal of the peroxide groups by reduction p revents enzyme inhibitio n. Such a mechanism is also supported by the inefficient inhibition of LDH by these peroxides, in the absence of metal ions, as this enzyme does not contain an active site thiol. The nonradical nature of the GAPDH and GR inhibitio n in the absence of metal ions is confirmed by the E PR studies where no r adicals were detected; previous studies have shown that radicals formed from these p eroxide can be detected using the methodology employed when peroxide formation is stimulated with metal ions [25,32,33]. Though it is possible that the inactivation of GAPDH and GR occurs via oxidation of nonactive site residues, that bring about loss of functional integrity, the stoichiometry of inactivation ( i.e. the loss of two molecules of thiol per molecule of peroxide consumed) suggests that the inactivating reaction(s) are highly specific. This is inconsistent with a radical-mediated process. This stoichi- ometry detected with GAPDH is consistent with the occurrence of both reaction 1 and subsequent reaction of the s ulfenic acid formed with a second thiol to give a disulfide bond (reaction 2). Previous studies have provided direct evidence for the formation of intramolecular disulfide bonds du ring oxidation of GAPDH between the active site thiol Cys149 and a further thiol residue, Cys153, which is in very close proximity to the former species [37,54,55]. No direct evidence for the formation of such a disulfide has been obtained in the current study, but such a mechanism seems highly likely on the basis of the data obtained. In contrast to this direct (nonradical) inactivation, inhibition of GR and LDH by H 2 O 2 is believed to occur via radical-mediated reactions, catalysed by the added Fe 2+ –EDTA. Enzyme-SH þ Peptide-/Protein-OOH ! Enzyme-S-OH þ Peptide-/Protein-OH ð1Þ Enzyme-S-OH þ Enzyme-SH ! Enzyme-S-S-Enzyme ð2Þ Previous studies have shown that some Met residues also react with hydroperoxides, to give the sulfoxide, with concomitant reduction of the hydroperoxide to the alcohol (e.g [56,57]). It is therefore possible that 1 O 2 -generated peptide and protein peroxides may also inhibit enzymes containing critical Met residues. These previous observa- tions are consistent with the statistically significant protec- tive effect offered by f ree Met in the i nhibition experiments carried out with N-Ac-Trp-OMe peroxides and GAPDH (cf. Table 2, P <0.01byone-way ANOVA with Dunnett’s posthoc test, when the Met treated sample was compared to photolysed N-Ac-Trp-OMe with added Fe 2+ –EDTA). However the extent of protection afforded by Met was much less marked that that seen with the thiol compounds and ascorbate at equimolar concentrations, suggesting that reaction of these peptide and protein peroxides with Met residues is kinetically uncompetitive when compared with reaction with activated (low pKa) Cys residues. A previous study has shown that peroxides formed by 1 O 2 on peptides and proteins a re not removed by catalase [25], and this has b een confirmed in the present study. These peroxides are also likely to be poor substrates for the glutathione peroxidase family, as a result of their steric bulk, and the buried position of most Tyr, His and Trp residues in proteins. This hypothesis is supported by a previous report that showed that radiation-generated protein peroxides are not removed rapidly by this enzyme, though some amino-acid peroxides are [27,29]. Reaction with low-molecular-mass reducing agents a nd antioxidants is therefore likely t o be t he major r oute for the removal of, or protection against, such peroxides in cells [26,29]. The studies reported here show that thiols can ameliorate inactivation of GAPDH induced by these 1 O 2 -generated peptide and protein peroxides, presumably by acting as sacrificial targets. This is in accord with the known rapid depletion of GSH and other thiols (both low-molecular- mass and protein-bound) in photo-oxidized cells, and that maintenance of thiol levels offers protection [58–61]. Similarly, it has been shown that ascorbate and thiols can readily remove radiation-generated peptide and p rotein peroxides [26,27,29]. It has also been sh own that over- expression, in human fibroblast cells, of the enzyme thioredoxin, which maintains low-molecular-mass thiols in a reduced form, protects cells against photo-oxidative damage and cell death [62,63]. Whether the p rotection offered by thiols is o wing to direct scavenging of 1 O 2 , removal of peroxides (H 2 O 2 and/or protein), or repair of reversibly damaged targets, such as the enzymes investi- gated here, r emains to be established. ACKNOWLEDGEMENTS The authors a re grateful to the Australian Research Council and the Juvenile Diabetes Foundation International for financial support, and to Dr Clare Hawkins for helpful discussions. REFERENCES 1. Rosen, H. & Michel, B.R. (1997) Redundant contribution of myeloperoxidase-dependent systems to neutrophil-mediated kill- ing of Escherichia coli. Infect. Immun. 65, 4173–4178. 2. Kiryu, C., Makiuchi, M., Miyazaki, J., Fujinaga, T. & Kakinuma, K. ( 1999) P hysiological production of singlet mole- Ó FEBS 2002 Enzyme inhibition by 1 O 2 -mediated protein peroxides (Eur. J. Biochem. 269) 1923 cular oxygen in the myeloperoxidase-H 2 O 2 -chloride system. FEBS Lett. 443, 154–158. 3. Kanofsky, J.R., Hoogland, H., Wever, R. & Weiss, S.J. (1988) Singlet o xygen production by human eosinophils. J. Biol. Chem. 20, 9692–9696. 4. Valenzeno, D.P. (1987) Photomodification of biological mem- branes with emphasis on singlet oxygen mechanisms. Photochem. Photobiol. 46, 147–160. 5. Teixeira, M.M., Cunha, F.Q., N oronha-Dutra, A. & Hothersall, J. (1999) Production of singlet oxygen by eosinophils activated in vitro by C5a and leukotriene B4. FEBS L ett. 453, 2 65–268. 6. Steinbeck, M.J., Khan, A.U. & Karnovsky, M.J. (1993) Extra- cellular production o f singlet oxygen by s timulated macrophages quantified using 9,10-diphenylanthracene and perylene in a poly- styrene film. J. Biol. C hem. 268, 15649–15854. 7. Sohal, R.S. & Weindruch, R. (1996) Oxidative stress, caloric restrict ion , and aging. Science 273, 59–63. 8. Black, H.S., deGruijl, F.R., Forbes, P.D., Cleaver, J.E., Anan- thaswamy, H.N., deFabo, E.C., Ullrich, S.E. & Tyrrell, R.M. (1997) Photocarcinogenesis: an overview. J. Photochem. Photobiol. B. 40, 29–47. 9. Dean, R.T., Fu, S., Stocker, R. & Davies, M.J. (1997) Biochem- istry and pathology of radical-mediated protein oxidation. Bio- chem. J. 324, 1–18. 10. Ortwerth, B.J., Casserly, T.A. & O lesen, P.R. (1998) Singlet oxy- gen production correlates with His and Trp destruction in brunescent cataract water-insoluble p roteins. Exp. Eye Res. 67, 377–380. 11. Stadtman, E .R. & Berlett, B.S. (1998) Reactive oxygen-mediated protein oxidation in aging and disease. Drug Metab. Rev. 30,225– 243. 12. Davies, M.J. & Truscott, R.J.W. (2001) Photo-oxidation o f pro- tein and its role in cataractogenesis. J. Photochem. Photobiol. B. 63, 114–125. 13. Lutgerink, J.T., van der Akker, E., Smeets, I., Pachen, D., van Dijk, P., Aubry, J M., Joenje, H., Lafleur, M.V.M. & Retel, J. (1992) Interaction of singlet oxygen with DNA and biological consequence s. Mutat. Res. 27 5, 377–386. 14. Cadet, J., Berger, M., Decarroz, C., Wagner, J.R., Van Lier, J.E., Ginot, Y.M. & Vigny, P. (1986) Photosensitised reactio ns of nucleic acids. Biochemie 68 , 813–834. 15. Vever-Bizet, C., Dellinger, M., Brault, D., Rougee, M. & Ben- sasson, R.V. (1989) Singlet molecular oxygen q uenching by satu- rated and unsaturated fatty-acids and by cholesterol. Photochem. Photobiol. 50, 321–325. 16. Girotti, A.W. (1992) Photosensitized oxidation of cholesterol in biological systems: reaction pathways, cytotoxic effects an d defense mechanisms. J. Photochem. Photobiol. B: Biol. 13 , 105–118. 17. Wagner,J.R.,Motchnik,P.A.,Stocker,R.,Sies,H.&Ames,B.N. (1993) The oxidation of blood plasma and low density lipoprotein components by chemically generated singlet oxygen. J. Biol. Chem. 268, 18502–18506. 18. Matsuo, I., Yoshino, K. & Ohkido, M. (1983) Mechanism of skin surface lipid peroxidation. Curr. Probl. Dermatol. 11, 135–143. 19. Straight, R.C. & Spikes, J.D. (1985) Photosensitized oxidation of biomolecules. In Singlet O 2 (Frimer, A.A., ed.), pp. 91–143. CRC Press, Boca Raton, FL, USA. 20. Bensasson, R.V., Land, E.J. & Truscott, T.G. (1993) Excited States and Free Radicals in Biology and Medicine. Oxford Uni- versity Press, Oxford, U K. 21. Wilkinson, F., Helman, W.P. & Ross, A.B. (1995) Rate constants for the decay and reactions o f the lowest el ectronically excited state of m olecular oxygen in solution. An expanded and revised com- pilation. J. Phys. Chem. Ref. Data. 24, 663–1021. 22. Saito, I., Matsuura, T., Nakagawa, M. & Hino, T. (1977) Per- oxidic intermediates in photosensitized oxygenation of tryptophan derivatives. Acc. Chem. Res. 10 , 346–352. 23. Jin, F .M., Leitich, J. & von Sonntag, C. (1995) The photolysis (k ¼ 254 nm) of tyrosine in aqueous solutions in the absence and presence of oxygen – the reaction of tyrosine with singlet oxygen. J. Photochem. Pho tobiol. A: Chem. 92, 147–153. 24. Tomita, M., Irie, M. & Ukita, T. (1969) Sensitized photooxidation of histidine and its derivatives. Products and mechanism of the reaction. Biochemist ry 8, 5149–5160. 25. Wright, A., Hawkins, C.L. & Davies, M.J. (2000) Singlet oxygen- mediated protein oxidation: evidence for the formation of reactive peroxides. Redox Report 5, 159–161. 26. Simpson, J.A., Narita, S., Gieseg, S., Gebicki, S., Gebicki, J.M. & Dean, R.T. (1992) Long-lived reactive species o n free-radical- damaged proteins. Biochem. J. 28 2 , 621–624. 27. Gebicki, J.M. (1997) Protein hydroperoxides as new reactive oxygen species. Redox Report 3, 99–110. 28. Hawkins, C.L. & Davies, M.J. (2001) Generation and propagation of radical reactions on proteins. Biochim. Biophys. Acta. 1504, 196–219. 29. Fu,S.,Gebicki,S.,Jessup,W.,Gebicki,J.M.&Dean,R.T.(1995) Biological fate of amino acid, peptide and protein hydroperoxides. Biochem. J. 311, 821–827. 30. Luxford, C., Morin, B., Dean, R.T. & Davies, M.J. (1999) Histone H1- and other protein- and amino acid-hydroperoxides can give rise to free radicals which oxidize DNA. Biochem. J. 344, 125–134. 31. Luxford, C., Dean, R.T. & Davies, M.J. (2000) Radicals derived from h istone hydroperoxides damage nucleobases in RNA and DNA. Chem. Res. T oxicol. 13, 665–672. 32. Davies, M.J., Fu, S. & Dean, R.T. (1995) Protein hydroper- oxides can give rise to reactive free radicals. Biochem. J. 305, 643–649. 33. Davies, M.J. (1996) Protein and peptide alkoxyl radicals can give rise to C-terminal decarboxylation and backbone cleavage. Arch. Biochem. Biophys. 336, 163–172. 34. Gebicki, S., Dean, R.T. & Gebicki, J.M. (1996) Inactivation of glutathione reductase by protein and amino acid peroxides. In Oxidative Stress and Redox Regulation: Cellular Signalling, AIDS, Cancer and Other Diseases. pp. 139. Institute Pasteur, Paris, France. 35. Soszynski, M., Filipiak, A., Bartosz, G. & Gebicki, J.M. (1996) Effect o f amino acid peroxides on the Erythrocyte. Free Radic. Biol. Med. 20, 45–51. 36. Gay, C., Collins, J. & Gebicki, J.M. (1999) Hydroperoxide assay with the ferric-xylenol orange comple x. Anal. Biochem. 273,149– 155. 37. Ishii, T., Sunami, O., Nakajima, H., Nishio, H., Takeuchi, T. & Hata, F. (1999) Critical role of sulfenic acid formation of thiols in the inactivation of glyceraldehyde-3-phosphate dehydrogenase by nitric oxide. Biochem. Pharmacol. 58, 133–143. 38. Silvester, J.A., Timmins, G.S. & Davies, M.J. (1998) Protein hydroperoxides and carbonyl groups generated by porphyrin- induced photo-oxid ation of bovine serum albumin. Arch. Biochem. Biophys. 350, 249–258. 39. Means, G.E. & Feeney, R.E. (1971) Chemical Modification of Proteins. Holden-Day, San Francisco, CA, USA. 40. Michaeli, A. & Feitelson, J. (1994) Reactivity of singlet oxygen towardaminoacidsandpeptides.Photochem. Photobiol. 59,284– 289. 41. Gieseg, S., Duggan, S. & Gebicki, J.M. (2000) Peroxidation of proteins before lipids in U937 cells exposed t o peroxyl radicals. Biochem. J. 350, 215–218. 42. Armstrong, D.A. & Buchanan, J.D. (1978) Reactions of O À 2 Æ, H 2 O 2 and other oxidants with sulfhydryl enzymes. Photochem. Photobiol. 28, 743–755. 43. Janero, D.R., Hreniuk, D. & Sharif, H.M. (1994) Hydroperoxide- induced oxidative stress impairs heart muscle cell carbohydrate metabolism. Am. J. Physiol. 266, C179–C188. 1924 P. E. Morgan et al. (Eur. J. Biochem. 269) Ó FEBS 2002 44. Ciolino, H.P. & Levine, R.L. (1997) Modification of p roteins in endothelial cell death during oxidative stress. Free Radic. Biol. Med. 22, 1277–1282. 45. Colussi, C., Albertini, M.C., Coppola, S., Rovidati, S., Galli, F. & Ghibelli, L . (2000) H 2 O 2 -induced block of glycolysis as an active ADP-ribosylation reaction protecting cells from apoptosis. FASEB J. 14, 2266–2276. 46. Hu, M.L. & Tappel, A.L. (1992) Potentiation of oxidative damage to proteins by ultraviolet-A and protection by antioxidants. Photochem. Photobiol. 56, 357–363. 47. Knight, K.L. & Mudd, J.B. (1984) The reaction of ozone with glyceraldehyde-3-phosphate dehydrogenase. Arch. Biochem. Bio- phys. 229, 259–269. 48. Van der Zee, J., Van Steveninck, J., K oster, J.F. & Dubbelman, T.M. (1989) Inhibition of e nzymes and o xidative damage of re d blood cells induced by t-butylhydroperoxide-derived radicals. Biochim. Biophys. Acta. 980, 175–180. 49. Buchanan, J.D. & Armstrong, D.A. ( 1978) The radiolysis o f gly- ceraldehyde-3-phosphate dehydrogenase. Int. J. Radiat. Biol. 33, 409–418. 50. Gutierrez-Co rrea, J. & Stoppani , A.O. (1997) Inactiva tion of yeast glutathione reductase by Fenton systems: effect of metal chelators, catecholamines and thiol compounds. Free Ra dic. Res. 27, 543– 555. 51. Tabatabaie, T. & Floyd, R.A. (1994) Susceptibility of glutathione peroxidase and glutathione reductase to oxidative damage and the protective effect o f spin t rapping agents. Arch. Biochem. Biophys. 314, 112–119. 52. Ogino, T. & Okada, S. (1995) Oxidative damage of bovine serum albumin and other enzyme proteins by iron-chelate complexe s. Biochim. Biophys. Acta. 1245, 359–365. 53. Dicker, E. & Cederbaum, A.I. (1988) Increased oxygen radical- dependent inactivation of metabolic enzymes by liver microsomes after chronic ethanol consumption. FASEB. J. 2, 2901–2906. 54. Ehring, R. & Colowick, S.P. (1969) The two-step formation and inactivation of acylphosphatase by agents acting on glyceraldehyde ph osphat e dehydrogenase. J. Biol. Chem. 244, 4589–4599. 55. Wassarman, P.M. & Major, J.P. (1969) The reactivity of the sulfhydryl groups of lobster muscle glyceraldehyde 3-phosphate dehydrogenase. Biochemistry 8, 1076–1082. 56. Garner, B., Witting, P.K., Waldeck, A.R., Christison, J.K., Raf- tery, M. & Stocker, R. (1998) Oxidation of high density lipopro- teins. I. Formation of methionine sulfoxide in apolipoproteins AI and AII is an early event that accompanies lipid peroxidation and can be enhanced by alpha-tocopherol. J. Biol. Chem. 27 3, 6080– 6087. 57. Garner, B., Waldeck, A .R., Witting, P.K., Rye,K A. &Stocker, R. (1998) Oxidation of high density lipoproteins. II. Evidence for direct reduction of lipid hydroperoxides by methionine residues of apolip op rot ein s AI and AII. J. Biol. Chem. 273, 6088–6095. 58. Truscott, R.J. & Augusteyn, R.C. (1977) The state of sulfydryl groups in normal and cataractous human lenses. Exp. Eye Res. 25, 139–148. 59. Garner, M.H. & Spector, A. (1980) Selective oxidation of cysteine and methionine in normal and senile cataractous lenses. Proc. Nat l Acad.Sci.USA77, 1274–1277. 60. Dillon, J., Roy, D. & Spector, A. (1985) The photolysis of lens fiber membranes. Exp. Eye R es. 41, 53–60. 61. Sweeney, M.H. & Truscott, R.J. (1998) An impediment to glutathione diffusion in older normal human lenses: a possible precondition for nuclear cataract. Ex p. Eye Res. 67, 587–595. 62. Didier, C., Pouget, J.P., Cadet, J., Favier, A., Beani, J.C. & Richard, M.J. (2001) Modulation o f e xogenous and endogenous levels of thioredoxin in human skin fibroblasts prevents DNA damaging effect of ultraviolet A radiation. Free Radic. Biol. Med. 30, 537–546. 63. Didier, C., Kerblat, I., Drouet, C., Favier, A., Beani, J. & Richard, M. (2001) Induction of thioredoxin by ultraviolet-A radiation prevents oxidative- mediated cell death in human skin fibroblasts. Free Radic. Biol. Med. 31, 585–598. Ó FEBS 2002 Enzyme inhibition by 1 O 2 -mediated protein peroxides (Eur. J. Biochem. 269) 1925 . Inhibition of glyceraldehyde-3-phosphate dehydrogenase by peptide and protein peroxides generated by singlet oxygen attack Philip E. Morgan 1 ,. peptides and proteins with singlet oxygen (generated by visible light in the presence of rose bengal dye) yields long-lived peptide and protein peroxides.