Superoxide radicals react with peptide-derived tryptophan radicals with very high rate constants to give hydroperoxides as major products

Author’s Accepted Manuscript Superoxide radicals react with peptide-derived tryptophan radicals with very high rate constants to give hydroperoxides as major products Luke Carroll, David I. Pattison, Justin B. Davies, Robert F. Anderson, Camilo Lopez-Alarcon, Michael J. Davies www.elsevier.com

PII: DOI: Reference:

S0891-5849(18)30090-X https://doi.org/10.1016/j.freeradbiomed.2018.02.033 FRB13642

To appear in: Free Radical Biology and Medicine Received date: 3 January 2018 Revised date: 22 February 2018 Accepted date: 23 February 2018 Cite this article as: Luke Carroll, David I. Pattison, Justin B. Davies, Robert F. Anderson, Camilo Lopez-Alarcon and Michael J. Davies, Superoxide radicals react with peptide-derived tryptophan radicals with very high rate constants to give hydroperoxides as major products, Free Radical Biology and Medicine, https://doi.org/10.1016/j.freeradbiomed.2018.02.033 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Superoxide radicals react with peptide-derived tryptophan radicals with very high rate constants to give hydroperoxides as major products Luke Carrolla,b,c, David I. Pattisona,b, Justin B. Daviesd, Robert F. Andersone, Camilo LopezAlarconf, Michael J. Daviesa,b,c* a

The Heart Research Institute, Sydney, Australia

b

Sydney Medical School, University of Sydney, Australia

c

Department of Biomedical Sciences, Panum Institute, University of Copenhagen, Denmark

d

Australian Nuclear Science and Technology Organisation, Lucas Heights, Australia

e

School of Chemical Sciences, University of Auckland, Auckland, New Zealand

f

Pontificia Universidad Catolica de Chile, Chile

*

Corresponding author at: Dept. of Biomedical Sciences, Building 12.6.30, Panum Institute,

University of Copenhagen, Blegdamsvej 3, Copenhagen 2200, Denmark. [email protected] (M.J. Davies)

ABSTRACT Oxidative damage is a common process in many biological systems and proteins are major targets for damage due to their high abundance and very high rate constants for reaction with many oxidants (both radicals and two-electron species). Tryptophan (Trp) residues on peptides and proteins are a major sink for a large range of biological oxidants as these sidechains have low radical reduction potentials. The resulting Trp-derived indolyl radicals (Trp●) have long lifetimes in some circumstances due to their delocalized structures, and undergo only slow reaction with molecular oxygen, unlike most other biological radicals. In contrast, we have shown previously that Trp● undergo rapid dimerization. In the current study, we

1

show that Trp● also undergo very fast reaction with superoxide radicals, O2●-, with k 1 – 2 x 109 M-1s-1. These values do not alter dramatically with peptide structure, but the values of k correlate with overall peptide positive charge, consistent with positive electrostatic interactions. These reactions compete favourably with Trp● dimerization and O2 addition, indicating that this may be a major fate in some circumstances. The Trp● + O2●- reactions occur primarily by addition, rather than electron transfer, with this resulting in high yields of Trp-derived hydroperoxides. Subsequent degradation of these species, both stimulated and native decay, gives rise to N-formylkynurenine, kynurenine, alcohols and diols. These data indicate that reaction of O2●-with Trp● should be considered as a major pathway to Trp degradation on peptides and proteins subjected to oxidative damage.

Keywords: Protein oxidation, Tryptophan, superoxide, hydroperoxide, dimerization, Nformylkynurenine, kynurenine

1. Introduction Under normal physiological conditions, free radicals are produced endogenously by cells via multiple metabolic (e.g. from mitochondria) and enzymatic reactions, and also as a component of the host immune defence system [1]. Free radicals are also produced in response to a wide variety of external stimuli, including radiation exposure and in response to toxins [1]. Cells have a range of antioxidants defence systems to mitigate and control the potential damage caused by scavenging or degrading oxidant species [1]. However, when these systems are overwhelmed, as a result of the inactivation or genetic loss of these systems, or as a result of excessive production of reactive species, damage can occur to

2

extracellular or cellular components which may contribute to cell and tissue dysfunction and disease [1-3]. Many reactive species, including both two-electron oxidants (e.g. HOCl, ONOOH, 1O2, O3) and free radicals, cause extensive damage to proteins, due to both the abundance of these species and their high rate constants for reaction with particular amino acid sidechains, with Cys, Met, cystine, Tyr, His and Trp being key targets due to their low radical reduction potentials [2,3]. Oxidation of these residues by radicals results in the formation of side-chain radicals (e.g. Cys-derived thiyl radicals, Tyr-derived phenoxyl radicals and Trpderived indolyl radicals (Trp●) [4-7]), which can then undergo further reactions [2,3]. These include addition reactions with O2 (to give peroxyl radicals, though these are typically slow and reversible with these stabilized radicals [2,3]), reaction with another (or similar) radical to give non-radical products (including dimers), hydrogen abstraction from compounds with weak X-H bond which can propagate damage, or -scission reactions that result in fragmentation of either the side-chains or backbone of the protein [2,3,8]. The superoxide radical-anion (O2●-) is produced by activated leukocytes (e.g. neutrophils, monocytes, and macrophages) as part of the immune response via the activity of the NADPH oxidase family of enzymes [9]. O2●- is also formed by most cells via electron leakage from the electron transfer pathways of mitochondria, the endoplasmic reticulum and the plasma membrane, with this process enhanced by organelle damage (e.g. mitochondrial dysfunction [10]). O2●- reacts slowly with most molecular targets, though it has been shown to disrupt iron-sulfur cluster enzymes [11,12]. However, O2●- can react rapidly with other radicals [12-16], to give other reactive species. Under physiological conditions, dismutation (either spontaneous or catalyzed by superoxide dismutases) of two O2●- to form H2O2 is a major reaction, with the latter capable of inducing further damage [17-19]. Reaction of O2●- with nitric oxide (NO●) gives the powerful oxidant peroxynitrite / peroxynitrous acid (ONOO-/ONOOH), that can react directly with biological targets, or undergoes limited homolysis to radicals that can induce oxidation and nitration (e.g. of proteins, lipids, DNA [13,14,20,21]). Thus, in spite of the low capacity of O2●- to induce direct damage to biological targets, its ability to form other reactive species has resulted in the suggestion that O2●- can be an important indirect mediator of damage to biological targets. O2●- can react rapidly with low-molecular-mass Tyr phenoxyl radicals, with the rate constant for such reactions, k, being ~ 1.5 × 109 M−1 s−1 [4,5,22-25]. Similar reactions have 3

been detected on intact proteins (k ~ 6 × 108 M−1 s−1, [22]). The major reaction pathway has been reported to be via addition, rather than electron transfer, with this resulting in the formation of tyrosine-derived peroxides [22,24-27]. Trp● and related species also appear to react rapidly with O2●- [5,28-30] [31-33]. The indolyl (nitrogen-centred) radical of N-acetyl-Trp-methyl ester reacts with O2●-, with k 1.2 × 109 M-1 s-1 at pH 10, to form a cyclic hydroperoxide product that decomposes to give products related to N-formylkynurenine (NFK) [28]. The latter can undergo subsequent hydrolysis to give kynurenine (Kyn) and downstream species [5,34]. Indole-derived natural products, such as serotonin and melatonin, can also be oxidized to indolyl radicals that react rapidly with O2●- [29,30]. These reactions yield diones, with the mechanism proposed to be via the formation of cyclic peroxides [29,30]. Although direct reaction of molecular O2 with Tyr phenoxyl radicals or Trp● is relatively slow (k < 105 M-1 s-1) [25,28,35], singlet oxygen, 1O2, undergoes rapid cycloaddition reactions with Tyr, His and Trp residues, to yield endoperoxides and then hydroperoxides on ring opening [28,36-41]. These hydroperoxides can modulate protein structure and function [2], as they preferentially target Cys residues, with this resulting in inactivation of a wide array of enzymes including caspases, glyceraldehyde dehydrogenase, sarco/endoplasmic reticulum Ca2+-ATPase and phosphatases [2,42-45]. The peroxides can also be reduced by metal ions to give alkoxyl radicals, and subsequently alkyl and peroxyl species, that can initiate further damage [2,36,46]. These data suggest that the formation of Trp●, and the subsequent reactions of these species with O2●- to give a hydroperoxide may be of biological significance. Furthermore, it is well established that Trp (and Tyr) residues can act as “sinks” for oxidizing species during electron-transfer reactions within proteins, due to their low radical reduction potentials [47-49]. Thus, initial damage at remote sites can be transferred within a protein structure to Trp and Tyr residues, with this enhancing the extent of oxidation of these residues, even if these are present at low abundance, and / or buried within a protein structure. In addition, the significant lifetimes of Trp● and Tyr-derived radicals (when compared to other species) may favor reaction with O2●-. However, the rate constants for the reaction of Trp● with O2●- in peptides, the role of structure and charge in modulating these reactions, the extent of reaction with O2●- compared to other processes such as dimerization, and the yield of any resulting hydroperoxides, remain poorly characterized. 4

This study therefore examined the hypothesis that reaction of peptide Trp● with O2●- would be a fast and efficient process, and that these reactions would give rise to high yields of hydroperoxides. These kinetic data have been compared with rate constants of other reactions of Trp●, such as dimer formation, and the nature and yields of the products of the O2●- reactions characterized.

2. Materials and methods 2.1.

Materials Trp-containing peptides were obtained from Bachem (Bubendorf, Switzerland), O2

was sourced from CoreGas (Yennora, Australia), acetonitrile was obtained from MerckMillipore (Ryde, Australia), pronase was obtained from Roche (Sydney, Australia), and d5-Trp was purchased from Cambridge Isotope Laboratories (Tewksbury, USA). All other reagents were from Sigma-Aldrich (Castle Hill, Australia). All solutions were prepared in sodium phosphate buffer (5 mM, pH 7.4), unless otherwise stated.

2.2.

Pulse radiolysis Trp-containing peptides (200 µM) were dissolved in sodium phosphate buffer (5

mM, pH 7.4) containing sodium azide (NaN3, 3 mM) and sodium formate (50 mM). Solutions were bubbled with O2 in closed vessels for 10 mins prior to pulse radiolysis measurements. Pulse radiolysis kinetic experiments were carried out at 22 ± 1 °C using the instrumentation at the University of Auckland's Free Radical Research Facility, which utilizes a 4-MeV linear accelerator to deliver 200 ns electron pulses with doses of 2 to 20 Gy to a 2 cm path-length optical cell. The optical detection system and dosimetry method have been described previously [50]. Reactions were monitored by optical spectroscopy at 510 nm. Under the experimental conditions employed, the initial water-derived transients formed (HO●, H●, eaq)

are rapidly converted to a mixture of Trp● (from reaction of HO● with N3- to give N3●,

which subsequently reacts selectively with Trp to give Trp● [51]) and O2●- (from quantitative reaction of e-aq and H● with O2, as well as decomposition of the formate radical formed upon reaction of the formate ion with HO●) as described previously [51,52] .

5

2.3.

Kinetic analysis Pseudo-first order analysis was used to determine rate constants for the reaction

between Trp● and O2●-. Initial concentrations of Trp● were determined by optical absorbance at 510 nm using the molar absorption coefficient, Ɛ = 1670 M-1 cm-1 [53], with initial absorbance determined by extrapolation of the initial decrease in absorbance at 510 nm to t = 0 s. O2●- concentrations were subsequently determined by subtracting the initial Trp● yield from the theoretical total radical yield (0.62 µM Gy-1). These conditions resulted in a > 5-fold excess of O2●- over Trp●. Observed rate constants, kobs, were determined by fitting single exponential curves to absorbance vs time plots obtained from pulse radiolysis, with the exponential fits carried out on residuals after subtracting the subsequent minor decay from the well-separated initial portion of the transients. The kobs values were plotted against the initial O2●- concentration (varied by altering the radiation dose), and the gradient of this line yielded the second order rate constant.

2.4. Steady state irradiation Trp-containing peptides (500 µM) were dissolved in phosphate buffer (pH 7.4, 5 mM) containing NaN3 (7.5 mM) and formate (125 mM). Solutions were bubbled with O2 for 10 min before, and then throughout the irradiation experiments. Samples were exposed to radiation from a Co60 source (GATRI facility, ANSTO, Lucas Heights) at either a low or high radiation flux (~ 18 or 40 Gy min-1) until the desired cumulative dose was reached (total radiation dose of 500 or 1000 Gy). Catalase (50 µL per 10 mL of sample; 1 mg mL-1) was added to each sample to remove any H2O2 (but not organic peroxides) formed during irradiation [36,40], and samples were aliquoted, frozen and stored until further analysis.

2.5.

Quantification of hydroperoxides The FOX assay was used to examine the presence of organic hydroperoxides in

irradiated samples. A working solution was prepared by mixing 1 part FeSO4 (25 mM) in H2SO4 (2.5 M) with 2 parts water and 1 part xylenol orange solution (10 mM in H2O), and the working solution was filtered with a 0.45 µm syringe filter. Ten µL of working solution was 6

pipetted into 96-well plates and 200 µL of previously irradiated samples (diluted 1 in 50 into water) added. The plate was gently mixed, incubated in the dark for 10 min, and the absorbance measured at 560 nm using a plate reader (M2e plate reader, Molecular Devices). A standard curve was constructed using 0 – 25 µM H2O2, and peroxide levels are reported as relative to H2O2 equivalents determined by comparison to this curve, and adjusted using a response factor based on the iodometric determination of hydroperoxides, carried out as reported previously [54,55]. Thus, 200 µL of H2O2 standards (0 – 100 µM) and GlyTrp samples were deoxygenated by bubbling with N2, before addition of 1 mL 5 % KI in methanol:acetic acid (2:1) solution that had also been deoxygenated by N2 bubbling. Samples were incubated for 15 min in the dark before addition of 100 µL cadmium acetate (8 % w/v) in 50 % MeOH at 4 oC. Triplicate samples (200 µL) were then placed in a 96-well plate and the absorbance at 358 nm measured using a plate reader (M2e plate reader, Molecular Devices). The resulting absolute hydroperoxide concentrations were used to determine a response factor for the FOX assay measurements as the stoichiometry of the Fe2+-xylenol orange reaction with hydroperoxides is not well established, and has been shown to be dependent on the nature of the hydroperoxide [55,56].

2.6.

Detection and quantification of Kyn and NFK formation in irradiated samples Irradiated samples, generated as described above, were spiked with d5-Trp as an

internal standard before being digested to constituent amino acid by incubation with pronase (0.1 mg mL-1) overnight at 37 °C. Pronase was removed by filtration through 3 kDa centrifugal devices (Pall). Samples were analysed by LC-MS as described below, and levels of Kyn and NFK were determined by comparison to standard curves prepared using authentic commercial materials.

2.7. LC-MS of irradiated samples Samples were prepared and filtered through 0.2 µm NanoSep centrifugal filter devices (Pall) for 2 min at 10,000 g prior to analysis. Samples were analysed on a LCQ Deca XP Max Plus mass spectrometer coupled to a Surveyor HPLC system (Thermo Electron Corp., Rydalmere, NSW, Australia). For mass spectrometric detection, the source voltage was set at 4.50 kV, and the capillary temperature was 250 °C. Nitrogen was used as both the sheath 7

(flow rate 60 arbitrary units) and sweep gas (flow rate 20 arbitrary units). For MS/MS, He was used as the collision gas for collision-induced dissociation (CID) experiments to generate fragments, with the normalized collision energy set to 35. Samples (10 µL) were injected onto a Zorbax ODS C-18 column (250 mm x 4.6 mm, 5 µm packing) and separated using a gradient elution method with a total flow rate of 0.4 mL min-1. Buffer A consisted 0.1 % trifluoroacetic acid (TFA) in water and buffer B consisted of 0.1 % TFA in 50 % acetonitrile (ACN) in water. Buffer B was held at 5 % for the first 10 min, before increasing to 75 % over the next 35 min and then held at 75 % B for 15 min, before decreasing back to 5 % over 5 min and re-equilibrating for 10 min. Under these conditions, Trp-containing peptides eluted between 35 and 60 min.

2.8.

Errors and statistics Kinetic data are presented as means, with 95 % confidence intervals, from n = 3

independent experiments. Data from product analyses are reported as mean and standard deviations (SD) from 6 determinations from 2 independent irradiations, unless stated otherwise. Statistical analyses were carried out in GraphPad Prism v 7.0, with p < 0.05 taken as significant.

3.

Results

3.1.

Determination of second-order rate constants for reaction of peptide-derived Trp●

with O2●O2-saturated solutions of Trp-containing peptides (200 µM) in phosphate buffer (5 mM, pH 7.4) containing NaN3 (3 mM) and formate (50 mM) were exposed to 200 ns electron pulses with varying total radiation doses, and the time-dependent decay of the generated Trp● monitored at 510 nm, which corresponds to max for Trp-derived indolyl radicals [51]. N3- and HCOO- compete for reaction with HO●, and at these concentrations, approximately 30 % of the HO● (total OH● yield of 0.28 µM Gy-1) react with N3- and subsequently form Trp● (rate constants [57]). The formate radical decomposes to yield O2●-, and O2 reacts with H● and e-aq to yield O2●-. The initial concentrations of Trp● and O2●- were calculated as described in the Materials and methods. Experimentally, the average Trp● yield was 0.11 µM Gy-1. The resulting kinetic traces obtained at 510 nm fitted well to single

8

exponential decay functions and allowed the determination of observed rate constants (kobs, Fig. 1A). The gradient of plots of kobs against the initial O2●- concentrations (varied by changing the radiation dose) yielded the second-order rate constants for reaction of the peptide-derived Trp● with O2●- (Fig. 1B). These values ranged between 7.6 x 108 and 2.2 x 109 M-1 s.1 for the peptides tested (Table 1) in line with previous more limited data [31,32]. The determined rate constants correlated in a positive manner with increasing overall peptide positive charge (Fig. 1C), indicating that a more positively-charged environment increases the rate at which O2●- reacts with Trp● (see also [32]).

3.2.

Reaction of Trp● and O2●- results in the formation of peroxides Previously irradiated samples, from which any radiation-generated H2O2 had been

removed by treatment with catalase, were assayed by the FOX assay for the presence of organic peroxides (Fig. 2). The FOX assay allows for the detection and relative quantification of peroxides by monitoring the absorbance change of the xylenol orange-Fe2+ complex upon oxidation by peroxides, to the Fe3+ species, with the absolute hydroperoxide concentrations determined by iodometric titration of the species detected with GlyTrp and ValTrp (Supplementary Fig. 1), and determination of a correction factor for the FOX assay (correction factor = 2.22). This was carried out as the stoichiometry of the xylenol orangeFe2+ / hydroperoxide reactions are poorly characterized [55]. Control, non-irradiated, peptide samples showed no reaction with the FOX reagent. Exposure of the Trp peptides to increasing total radiation doses (500 or 1000 Gy) resulted in statistically-significant increases in peroxide levels for all the peptides tested (Fig. 2). For the 1000 Gy dose, two different dose rates were also investigated, but no significant differences were detected between the peroxide levels detected with these different dose rates. In each case, the highest radiation dose used (1000 Gy) generated organic peroxides representing 93-125% of the total yield of Trp●.

3.3.

Detection and quantification of peptide-derived NFK and Kyn in samples immediately

after irradiation Organic peroxides formed on free Trp and simple Trp-derivatives have been shown previously to undergo decomposition under a range of different conditions, with this 9

resulting in the formation of ring-opened products from the indole ring. Thus peroxide degradation gives NFK, and subsequently Kyn by hydrolysis [2,36,58,59]. In order to test the significance of this pathway with Trp-containing peptides, previously irradiated (1000 Gy cumulative dose) and control peptides samples were digested overnight at 37 °C using pronase to give the free amino acids (and products therefrom), before removal of pronase by filtration through 3 kDa cut-off filters, and analysis of the filtrates by LC-MS. NFK was detected at concentrations of less than 1 µM in all of the control, nonirradiated samples. For irradiated GluTrp, GlyTrp, LysTrp and ValTrp, the concentrations of NFK detected were ~ 17 µM (Fig. 3C), representing a ~ 3 % conversion of the initial total radical flux to NFK (or ~ 15 % of Trp●). In contrast, LysTrpLys and GlyTrpGly yielded lower concentrations of NFK (~ 4 µM). In the latter case the poor solubility of this material may contribute to this low yield. Kyn was not detected in any of the control samples, but was detected at low levels (0.2 – 0.4 µM) in all of the irradiated peptide samples analysed (Fig. 3D). This represents a yield of less than 0.1 % of the total radical formation.

3.3.

Characterization of oxidation products arising from irradiation and subsequent

forced peroxide decomposition Organic peroxides can decompose in the presence of heat, light, metal ions and reductants to yield a range of products, including alcohols, diols and carbonyls. The yield of each of these species has been shown to be highly dependent on the species that induces decomposition [2,36,58,59]. As a consequence, the formation of such products, which might be formed from the Trp-containing peptides under forced decomposition conditions, was examined. Multiple oxidized products were detected in each irradiated sample, under the different decomposition conditions examined, with alcohols, Kyn and NFK species being the primary products (see Table 2). Kyn peptide derivatives were characterized by ions with m/z +4 compared to the parent peptide, with the peak areas of these ions markedly increased in the samples upon irradiation when compared to control peptides. In samples which were incubated overnight at 37 oC, following irradiation, the peak areas of the ions assigned to Kyn increased, probably due to the hydrolysis of NFK, and decreased upon NaBH4 treatment due to reduction of Kyn (see Fig. 4B for GlyTrp).

10

A number of alcohols (with m/z +16 compared to the parent peptide) were detected in each sample (e.g. 2 peaks were detected from irradiated GlyTrp at retention times 5.9 min and 7.4 min), and this is believed to reflect the potential formation of multiple isomers upon Trp oxidation. The alcohol peak for GlyTrp that eluted at 5.9 min increased in intensity after irradiation, but this did not increase further with overnight incubation at 37 oC. In contrast, NaBH4 treatment resulted in an increase in peak intensity, indicating that this species is likely to arise from reduction of a peroxide or carbonyl (Fig. 4C). Similarly, the peak detected at 7.4 min from GlyTrp increased in intensity on NaBH4 treatment of the irradiated sample, though this species was not detected in control or irradiated samples that were not treated with NaBH4 (Fig. 4D). NFK derivatives (detected as peaks with m/z +32 compared to the parent peptide) were the major product detected (in terms of relative peak areas, with these being at least 2 orders of magnitude more abundant than other products) in the irradiated samples of GluTrp, GlyTrp and ValTrp. Characterization of these peaks by MS/MS showed a loss of 42 mass units that is characteristic of NFK [36,38]. The intensity of the peaks from the NFK species were considerably higher in the irradiated samples than (non-irradiated) controls (Fig. 4A for GlyTrp). The intensity of these peaks decreased with overnight incubation, consistent with decomposition (probably hydrolysis), and a more significant decrease was observed with NaBH4 treatment, indicating that this species (or its precursor) are susceptible to reduction. Peaks (with m/z + 32 compared to the parent) that behave in a similar way to NFK were detected in KW and KWK irradiated samples, however, the MS/MS data for these peaks demonstrated an m/z loss of 111 and 110 respectively. As peroxides also react in a similar way to NFK, though typically fragment with a loss of m/z 17 or 34, these species have not been conclusively identified.

4.

DISCUSSION Radiolysis of water results in the initial formation of HO●, H● and e-eq. Under the reaction

conditions employed in these studies for pulse radiolysis and steady-state irradiation (3 : 50 ratio of NaN3 : formate; Reactions 1-4), ~ 30 % of the HO● generated will react with the N3to give N3● (Reaction 6), which selectively oxidize Trp side-chains to Trp● (Reaction 7) with a theoretical yield of 0.09 µM Gy-1. The remaining primary radicals and e-eq react either 11

directly with O2 to yield O2●- (Reaction 2), or formate to give the formate radical-anion (Reaction 4) which undergoes subsequent reaction with O2 to give O2●- (Reaction 5), with a theoretical yield of 0.533 µM Gy-1 superoxide. The concentrations of O2●- were determined by subtracting the Trp● concentration calculated from A0, from the total radical yield known to be generated on radiolysis of water (0.62 µM Gy-1). The experimentally observed yields of Trp● was 0.11 µM Gy-1 with an observed excess of O2●- over Trp● of ~5-fold; this excess yielded good pseudo first order plots to be obtained. (1)

→

(2)

→ →

(3)

↔

(4)

→

(5)

→

(6)

→

(7)

→

This study demonstrates that Trp● formed on multiple Trp-containing peptides, react with O2●- with rate constants in the range of k 0.8 – 2.2 x 109 M-1 s-1 (Table 1, Fig. 5) at physiological pH (7.4). This is consistent with a previous report on the rate constant for reaction of Trp● from N-acetyltryptophan methyl ester with O2●- of 1.2 x 109 M-1 s-1 at high pH (pH 10; [28]) and 2.3 x 109 M-1 s-1 for free Trp● [31].These values do not show marked variation with peptide structure, but a positive correlation has been demonstrated between the rate constant for reaction and the overall peptide positive charge (see also [32]). Thus, local concentrations of positive charge appear to enhance the rate of reaction, as might be expected from favorable electrostatic interactions. These rate constants are the highest known rate constants for Trp●, suggesting that such reactions may be competitive processes in vivo, if relatively high concentrations of each radical are present. It has been shown that Trp● radicals can have considerable (millisecondsecond) lifetimes in some proteins, suggesting that relatively high levels of these radicals can be formed. The overall significance of these reactions may therefore be limited by the concentration of O2●-, due to the rapid disproportionation of this radical, via both 12

spontaneous and enzyme- (SOD) catalyzed reactions. It has been reported that steady-state concentrations of O2●- in biological systems are in the nano- to pico-molar range [60], however much higher concentrations (in the micromolar range) can be present in, for example, the phagolysosomal compartments of neutrophils and other leukocytes upon activation due to the high catalytic activity of NADPH oxidase (NOx) enzymes [61]. It should however be noted that the rate constants for reaction of Trp ● with O2●- determined here, compare very favourably with the rate constants for both spontaneous (k 9.7 x 107 M-1 s-1) and SOD-catalysed reactions (k ~ 2 x 109 M-1 s-1 ) [60]. Consistent with the rapid reaction of O2●- with Trp●, high yields of organic peroxides were detected in all of the irradiated samples as determined by both the FOX assay and iodometric titration (Table 3, Fig. 5). The detection of these species is consistent with a mechanism involving addition of O2●- to the indole ring, rather than electron transfer, which would give the parent indole and O2 as products. This conclusion is consistent with some (but not all) previous studies (see also below) [28]. The relative levels of peroxides formed in these studies has been determined using the technically-simple FOX assay, with the absolute levels determined using the more challenging iodometric assay. The use of the latter has allowed a correction factor to be determined for the FOX assay, as the stoichiometry of the latter assay is poorly characterized [55]. The resulting factor is in line with some previously reported values, where a similar approach was adopted [55]. The same correction factor was used for all the materials examined as the peroxide(s) present in each case are likely to be Trp-derived, and similar in structure. Peroxides were detected for all of the peptides examined, consistent with a common process, and the levels of these materials increased with the overall radiation dose applied (Fig. 2). The observation that the two different dose rates examined (~ 18 and 40 Gy min-1) did not alter the overall yield of peroxides, is consistent with the addition of O2●- to Trp● being the major reaction, even when the steady-state concentrations of these species are varied significantly. The highest dose of radiation employed (1000 Gy) resulted in the detection of 157 – 211 µM organic peroxide. This gives a yield of 93-125% based on a Trp● concentration of ~ 168 µM formed at this dose (Table 3). The mechanism by which small excesses of peroxides are formed in excess over the Trp● formed by N3● is unknown, though there are a number of potential avenues for reaction. Peroxyl radicals can initiate chain reactions, undergoing hydrogen abstraction [2], potentially forming a secondary Trp● which 13

can in turn react with the excess O2●-. Alternatively, formation of backbone radicals by reaction with OH● and H● may also react with O2●-, yielding other (non-Trp derived) peroxides [2]. Scavenging of radicals by the backbone would decrease the calculated O2●yield reported in the experiments, and while this should not have a significant impact on the determined rate, the reported values may be underestimated if the yield of O2●- is decreased. In more complex systems, other reactions (e.g. repair of Trp● by reducing agents, such as ascorbate and GSH [62,63]) may limit this reaction and hence decrease the efficiency of hydroperoxide formation. In addition to the detection of hydroperoxides as a major initial product, significant levels of NFK were also detected, and the levels of this material can be increased by peroxide degradation (Table 3, Fig. 5). Under similar conditions to those employed in the current study, Fang et al [28] observed formation of both organic peroxides and NFKderivatives upon irradiation of N-acetyl-O-methyltryptophan, though with a higher ratio of NFK-derivatives to peroxide than observed here. This difference may be due to the conditions employed, as NFK is a known degradation product of Trp-derived peroxides [28,34,36], and hence an increased extent of peroxide degradation before analysis may account for this different ratio of products. In contrast to the reasonable agreement between the current study and that of Fang et al [28] and Santus et al [31] which have both employed radiation chemistry methods, previous studies [64,65] [31,32] have suggested that the electron transfer from O2●- to Trp● is the predominant pathway, rather than addition. While this pathway is thermodynamically possible, standard reduction potentials of Trp● (1.0 V) and O2●- (0.94 V) would suggest it is an unlikely pathway [65-67], and the high yields of peroxides detected here support the conclusion that addition is the major pathway. There are very significant experimental differences between the system employed here, and these previous studies which have employed light and the sensitizers, pterin and kynurenic acid, to generate Trp● and O2●- [64,65]. These systems therefore contain multiple other transient and reactive species (singlet and triplet states and other radicals) which may result in competing reactions and / or light-induced peroxide decomposition. Trp● are known to undergo rapid dimerization reactions to give (multiple isomeric) cross-linked species [52,65] (Fig. 5). These termination reactions, though fast, have significantly slower rate constants (~ 5-10-fold lower, k 0.6 – 2 x 108 M-1 s-1; see Table 1) [28,52]. Under the conditions used in the current study, where an excess of O2●- was 14

present, dimerization should not be favoured, and this is consistent with the high yield of peroxides and the absence of any (detectable) dimeric species. In biological situations where high fluxes of O2●- are formed (e.g. at sites of inflammation, where neutrophils and other immune cells can produce large amounts of O2●- [10]), this may also be true. However, Trp● dimerization may predominate in tissues with low steady-state O2 levels, or where the O2●- flux is low (e.g. in diseased or hypoxic/anoxic tissues, and during normal metabolism of some tissues with poor oxygen perfusion such as the centre of the lens). Thus, the formation of Trp dimers may be most significant at, and a marker of, low levels of O2●- formation, tissue hypoxia and tissue damage. These reactions of Trp● show some analogies to those of Tyr phenoxyl radicals, (Tyr●) which also react with O2●- to form hydroperoxides (e.g. [22,26,27]. The presence of SOD has been shown to inhibit the formation of peroxides formed on Tyr by O2●-, and to promote the formation of diTyr [22,26,68], suggesting that, as with Trp, reaction with O2●- competes with Tyr● dimerization to give the well-established cross-link, di-tyrosine. It should also be borne in mind that a contribution to the overall peroxide yields that might be detected in more complex systems, might also come from addition of O2 to Tyr● and Trp● to give peroxyl radicals and subsequently hydroperoxides by hydrogen atom abstractions. However, for both Tyr● and Trp● these O2 addition reactions are slow (for Tyr k << 105 M-1 s-1 and for Trp k < 105 M-1 s-1 [25,69]) and this does not appear to influence the O2●- reaction [28]. The mass spectrometric analysis carried out on the reaction products has allowed the detection of peaks with m/z +32 relative to the parent peptide, which would be consistent with peroxide formation, however no fragmentation spectra could be obtained that could conclusively characterize such species [36]. It is possible that this is due to facile peroxide decomposition under the mass spectrometry conditions with fragmentation prior to analysis, or poor ionization of the samples leading to weak signals. In contrast to this failure to unequivocably detect peroxide species by MS analysis, strong evidence was obtained in many of the samples (e.g. GluTrp, GlyTrp and ValTrp) for the presence of NFK which also gives rise to m/z +32 ions, with these assignments made on the basis of the characteristic loss of m/z 42 under MS/MS conditions. The detection of Kyn and NFK by LCMS after enzymatic digestion of all irradiated peptides (Fig. 3), confirms the formation of these species, and also makes it likely that the m/z +32 peak detected for LysTrpLys and LysTrp is also due to NFK formation on these peptides. Kyn-containing peptides and isolated 15

kynurenine was also detected in many of the samples, though at much lower concentrations. Both NFK and Kyn are well-characterized oxidation products of Trp, and have been detected, by various approaches, after treatment of Trp containing peptides and proteins with by a wide variety of oxidants including 1O2, H2O2 in the presence of Fe2+, and radiation amongst others [5,34,36,58,70-73]. Elevated levels of Kyn and NFK in tissue samples has been shown to correlate with tissue damage in a number of diseases [2,74], with, for example, NFK detected in the lens of bovine eyes after exposure to oxidation and radiation, with the presence of this species being associated with increased crystallin protein aggregation and cataract formation [75,76]. It should however, be noted that Kyn and NFK are also produced enzymatically via the action of tryptophan dioxygenase and indoleamine-2,3- dioxygenase enzymes [5,77,78], and hence may not be unequivocable markers of radical or oxidant formation. In contrast, there are no enzymatic pathways that we are aware of that generate Trp and Tyr dimers that do not involve free radical formation, though the radicals may initially arise via enzymatic activity. The detection of multiple other products with m/z +16 and +32 compared to the parent peptide, by MS analysis (Table 2), indicates that the decomposition chemistry of Trp is complex and involves multiple competing reactions (Fig. 5). This is consistent with previous studies on these and related species [36,79-81]. These products have been characterized as alcohols or diols based on the relative abundance remaining after incubation at 37 oC over extended periods, or NaBH4 treatment, consistent with these earlier reports. Alcohols are known products of multiple pathways of peroxide degradation, including direct 2-electron reduction, peroxide bond homolysis, and one-electron reduction to alkoxyl radicals with subsequent hydrogen abstraction from surrounding protein. The detection of these materials is therefore consistent with the major reaction pathway occurring in the current studies being addition of O2●- to Trp● to give a peroxide and subsequent degradation of these species [2,36,58,59]. Overall, this study demonstrates that Trp● reacts very rapidly with O2●- with second order rate constants, k, in the range 0.7 - 2.2 x 109 M-1 s.1, with the major initial products being hydroperoxides, with these being formed in near stoichiometric yields. Subsequent decomposition of these species gives NFK as a major initial product, and subsequently alcohol, diol and hydrolysis (Kyn) products. These reactions may be important fates of Trp●

16

formed in the presence of O2●-, with an alternative pathway for Trp● being dimerization under conditions where O2●- is absent, or present only at low concentrations.

Acknowledgements The authors thank Roger van Ryn at Auckland University for technical assistance with pulse radiolysis studies, and Sohil Sheth, Connie Banos and Allan Perry for technical assistance with steady-state irradiation studies. These studies were supported by grants from the Australian Research Council (DP140103116), the Australian Institute of Nuclear Science and Engineering (ALNGRA15521 and ALNGRA14021), the Novo Nordisk Foundation (grant NNF13OC0004294 to MJD) and FONDECYT grant (1141142 to CLA).

REFERENCES [1]

Halliwell, B.; Gutteridge, J. M. C. Free radicals in biology & medicine. Oxford: Oxford

University Press; 2015. [2]

Davies, Michael J. Protein oxidation and peroxidation. Biochem. J. 473 (2016) 805-

825. [3]

Hawkins, C. L.; Davies, M. J. Generation and propagation of radical reactions on

proteins. Biochim. Biophys. Acta-Bioenergetics 1504 (2001) 196-219. [4]

Houee-Levin, C.; Bobrowski, K.; Horakova, L.; Karademir, B.; Schoneich, C.; Davies, M.

J.; Spickett, C. M. Exploring oxidative modifications of tyrosine: An update on mechanisms of formation, advances in analysis and biological consequences. Free Radic. Res. 49 (2015) 347373. [5]

Ehrenshaft, M.; Deterding, L. J.; Mason, R. P. Tripping up trp: Modification of protein

tryptophan residues by reactive oxygen species, modes of detection, and biological consequences. Free Radic. Biol. Med. 89 (2015) 220-228.

17

[6]

Trujillo, M.; Alvarez, B.; Radi, R. One- and two-electron oxidation of thiols:

Mechanisms, kinetics and biological fates. Free Radic. Res. 50 (2016) 150-171. [7]

Schoneich, C. Thiyl radicals and induction of protein degradation. Free Radic. Res. 50

(2016) 143-149. [8]

Headlam, H. A.; Davies, M. J. Markers of protein oxidation: Different oxidants give

rise to variable yields of bound and released carbonyl products. Free Radic. Biol. Med. 36 (2004) 1175-1184. [9]

Babior, B. M. The respiratory burst oxidase. TIBS 12 (1987) 241-243.

[10]

Hayyan, M.; Hashim, M. A.; AlNashef, I. M. Superoxide ion: Generation and chemical

implications. Chem. Rev. 116 (2016) 3029-3085. [11]

Liochev, S. I.; Fridovich, I. Superoxide and iron: Partners in crime. IUBMB Life 48

(1999) 157-161. [12]

Winterbourn, C. C.; Kettle, A. J. Radical-radical reactions of superoxide: A potential

route to toxicity. Biochem. Biophys. Res. Commun. 305 (2003) 729-736. [13]

Huie, R. E.; Padmaja, S. The reaction of NO with superoxide. Free Radic. Res.

Commun. 18 (1993) 195-199. [14]

Goldstein, S.; Czapski, G.; Lind, J.; Merényi, G. Tyrosine nitration by simultaneous

generation of ⋅NO and O2⨪ under physiological conditions: How the radicals do the job. J. Biol. Chem. 275 (2000) 3031-3036. [15]

Jonsson, M.; Lind, J.; Reitberger, T.; Eriksen, T. E.; Merenyi, G. Free radical

combination reactions involving phenoxyl radicals. J. Phys. Chem. 97 (1993) 8229-8233. [16]

Cadenas, E.; Merényi, G.; Lind, J. Pulse radiolysis study on the reactivity of trolox c

phenoxyl radical with superoxide anion. FEBS Lett. 253 (1989) 235-238. [17]

Fridovich, I. Superoxide dismutases. Annu. Rev. Biochem. 44 (1975) 147-159.

18

[18]

Sies, H. Hydrogen peroxide as a central redox signaling molecule in physiological

oxidative stress: Oxidative eustress. Redox Biol. 11 (2017) 613-619. [19]

Winterbourn, C. C. Biological production, detection, and fate of hydrogen peroxide.

Antioxid. Redox Signal. (2017) doi: 10.1089/ars.2017.7425. [Epub ahead of print] [20]

Radi, R. Peroxynitrite, a stealthy biological oxidant. J. Biol. Chem. 288 (2013) 26464-

26472. [21]

Koppenol, W. H.; Bounds, P. L.; Nauser, T.; Kissner, R.; Rüegger, H. Peroxynitrous

acid: Controversy and consensus surrounding an enigmatic oxidant. Dalton Trans. 41 (2012) 13779-13787. [22]

Das, A. B.; Nauser, T.; Koppenol, W. H.; Kettle, A. J.; Winterbourn, C. C.; Nagy, P.

Rapid reaction of superoxide with insulin-tyrosyl radicals to generate a hydroperoxide with subsequent glutathione addition. Free Radic. Biol. Med. 70 (2014) 86-95. [23]

Nagy, P.; Kettle, A. J.; Winterbourn, C. C. Superoxide-mediated formation of tyrosine

hydroperoxides and methionine sulfoxide in peptides through radical addition and intramolecular oxygen transfer. J. Biol. Chem. 284 (2009) 14723-14733. [24]

Jin, F.; Leitich, J.; von Sonntag, C. The superoxide radical reacts with tyrosine-derived

phenoxyl radicals by addition rather than by electron transfer. J. Chem. Soc., Perkin Trans. 2 (1993) 1583-1588. [25]

Hunter, E. P. L.; Desrosiers, M. F.; Simic, M. G. The effect of oxygen, antioxidants, and

superoxide radical on tyrosine phenoxyl radical dimerization. Free Radic. Biol. Med. 6 (1989) 581-585. [26]

Winterbourn, C. C.; Parsons-Mair, H. N.; Gebicki, S.; Gebicki, J. M.; Davies, M. J.

Requirements for superoxide-dependent tyrosine hydroperoxide formation in peptides. Biochem. J. 381 (2004) 241-248.

19

[27]

Moller, M. N.; Hatch, D. M.; Kim, H. Y.; Porter, N. A. Superoxide reaction with tyrosyl

radicals generates para-hydroperoxy and para-hydroxy derivatives of tyrosine. J. Am. Chem. Soc. 134 (2012) 16773-16780. [28]

Fang, X.; Jin, F.; Jin, H.; von Sonntag, C. Reaction of the superoxide radical with the n-

centered radical derived from N-acetyltryptophan methyl ester. J. Chem. Soc., Perkin Trans. 2 (1998) 259-264. [29]

Wrona, M. Z.; Dryhurst, G. Oxidation of serotonin by superoxide radical: Implications

to neurodegenerative brain disorders. Chem. Res. Toxicol. 11 (1998) 639-650. [30]

Silva, S. O.; Ximenes, V. F.; Catalani, L. H.; Campa, A. Myeloperoxidase-catalyzed

oxidation of melatonin by activated neutrophils. Biochem. Biophys. Res. Commun. 279 (2000) 657-662. [31]

Santus, R.; Patterson, L. K.; Bazin, M. The diffusion-controlled reaction of

semioxidized tryptophan with the superoxide radical anion. Free Radic. Biol. Med. 19 (1995) 837-842. [32]

Santus, R.; Patterson, L. K.; Bazin, M.; Maziere, J. C.; Morliere, P. Intra and

intermolecular charge effects on the reaction of the superoxide radical anion with semioxidized tryptophan in peptides and N-acetyl tryptophan. Free Radic. Res. 29 (1998) 409419. [33]

Santus, R.; Patterson, L. K.; Hug, G. L.; Bazin, M.; Maziere, J. C.; Morliere, P.

Interactions of superoxide anion with enzyme radicals: Kinetics of reaction with lysozyme tryptophan radicals and corresponding effects on tyrosine electron transfer. Free Radic. Res. 33 (2000) 383-391. [34]

Saito, I.; Matsuura, T.; Nakagawa, M.; Hino, T. Peroxidic intermediates in

photosensitized oxygenation of tryptophan derivatives. Acc. Chem. Res. 10 (1977) 346-352.

20

[35]

Gebicki, S.; Gebicki, J. M. Formation of peroxides in amino acids and proteins

exposed to oxygen free radicals. Biochem. J. 289 (1993) 743-749. [36]

Gracanin, M.; Hawkins, C. L.; Pattison, D. I.; Davies, M. J. Singlet-oxygen-mediated

amino acid and protein oxidation: Formation of tryptophan peroxides and decomposition products. Free Radic. Biol. Med. 47 (2009) 92-102. [37]

Ronsein, G. E.; Bof de Oliveira, M. C.; Medeiros, M. H. G.; Di Mascio, P.

Characterization of O2 ((1)delta(g))-derived oxidation products of tryptophan: A combination of tandem mass spectrometry analyses and isotopic labeling studies. J. Am. Soc. Mass Spectrom. 20 (2009) 188-197. [38]

Ronsein, G. E.; Oliveira, M. C. B.; Miyamoto, S.; Medeiros, M. H. G.; Di Mascio, P.

Tryptophan oxidation by singlet molecular oxygen O2 ((1)delta(g)) : Mechanistic studies using O-18-labeled hydroperoxides, mass spectrometry, and light emission measurements. Chem. Res. Toxicol. 21 (2008) 1271-1283. [39]

Wright, A.; Bubb, W. A.; Hawkins, C. L.; Davies, M. J. Singlet oxygen-mediated protein

oxidation: Evidence for the formation of reactive side chain peroxides on tyrosine residues. Photochem. Photobiol. 76 (2002) 35-46. [40]

Agon, V. V.; Bubb, W. A.; Wright, A.; Hawkins, C. L.; Davies, M. J. Sensitizer-mediated

photooxidation of histidine residues: Evidence for the formation of reactive side-chain peroxides. Free Radic. Biol. Med. 40 (2006) 698-710. [41]

Jin, F.; Leitich, J.; von Sonntag, C. The photolysis (λ = 254 nm) of tyrosine in aqueous

solutions in the absence and presence of oxygen. The reaction of tyrosine with singlet oxygen. J. Photochem. Photobiol. A: Chemistry 92 (1995) 147-153. [42]

Hampton, M. B.; Morgan, P. E.; Davies, M. J. Inactivation of cellular caspases by

peptide-derived tryptophan and tyrosine peroxides. FEBS Lett. 527 (2002) 289-292.

21

[43]

Morgan, P. E.; Dean, R. T.; Davies, M. J. Inhibition of glyceraldehyde-3-phosphate

dehydrogenase by peptide and protein peroxides generated by singlet oxygen attack. Eur. J. Biochem. 269 (2002) 1916-1925. [44]

Dremina, E. S.; Sharov, V. S.; Davies, M. J.; Schoneich, C. Oxidation and inactivation of

serca by selective reaction of cysteine residues with amino acid peroxides. Chem. Res. Toxicol. 20 (2007) 1462-1469. [45]

Gracanin, M.; Davies, M. J. Inhibition of protein tyrosine phosphatases by amino

acid, peptide, and protein hydroperoxides: Potential modulation of cell signaling by protein oxidation products. Free Radic. Biol. Med. 42 (2007) 1543-1551. [46]

Davies, M. J.; Fu, S. L.; Dean, R. T. Protein hydroperoxides can give rise to reactive

free-radicals. Biochem. J. 305 (1995) 643-649. [47]

Prutz, W. A.; Butler, J.; Land, E. J.; Swallow, A. J. Direct demonstration of electron

transfer between tryptophan and tyrosine in proteins. Biochem. Biophys. Res. Commun. 96 (1980) 408-414. [48]

Prutz, W. A.; Butler, J.; Land, E. J.; Swallow, A. J. Unpaired electron migration

between aromatic and sulfur peptide units. Free Radic. Res. Commun. 2 (1986) 69-75. [49]

Faraggi, M.; DeFelippis, M. R.; Klapper, M. H. Long-range electron transfer between

tyrosine and tryptophan in peptides. J. Am. Chem. Soc. 111 (1989) 5141-5145. [50]

Anderson, R. F.; Denny, W. A.; Li, W. J.; Packer, J. E.; Tercel, M.; Wilson, W. R. Pulse

radiolysis studies on the fragmentation of arylmethyl quaternary nitrogen mustards by oneelectron reduction in aqueous solution. J. Phys. Chem. A 101 (1997) 9704-9709. [51]

Land, E. J.; Prutz, W. A. Fast one-electron oxidation of tryptophan by azide radicals.

Int. J. Radiat. Biol. 32 (1977) 203-207.

22

[52]

Carroll, L.; Pattison, D. I.; Davies, J. B.; Anderson, R. F.; Lopez-Alarcon, C.; Davies, M.

J. Formation and detection of oxidant-generated tryptophan dimers in peptides and proteins. Free Radic. Biol. Med. 113 (2017) 132-142. [53]

Prutz, W. A.; Land, E. J. Charge-transfer in peptides - pulse-radiolysis investigation of

one-electron reactions in dipeptides of tryptophan and tyrosine. Int. J. Radiat. Biol. 36 (1979) 513-520. [54]

Jessup, W.; Dean, R. T.; Gebicki, J. M. Iodometric determination of hydroperoxides in

lipids and proteins. Meth. Enzymol. 233 (1994) 289-303. [55]

Gay, C.; Collins, J.; Gebicki, J. M. Hydroperoxide assay with the ferric-xylenol orange

complex. Anal. Biochem. 273 (1999) 149-155. [56]

Bou, R.; Codony, R.; Tres, A.; Decker, E. A.; Guardicila, F. Determination of

hydroperoxides in foods and biological samples by the ferrous oxidation-xylenol orange method: A review of the factors that influence the method's performance. Anal. Biochem. 377 (2008) 1-15. [57]

Buxton, G. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. B. Critical review of rate

constants for reactions of hydrated electrons, hydrogen atoms and hydroxyl radicals (⋅OH/⋅O− in aqueous solution. J. Phys. Chem. Ref. Data 17 (1988) 513-886. [58]

Domingues, M. R.; Domingues, P.; Reis, A.; Fonseca, C.; Amado, F. M.; Ferrer-Correia,

A. J. Identification of oxidation products and free radicals of tryptophan by mass spectrometry. J. Am. Soc. Mass Spectrom. 14 (2003) 406-416. [59]

Simat, T. J.; Steinhart, H. Oxidation of free tryptophan and tryptophan residues in

peptides and proteins. J. Agric. Food Chem. 46 (1998) 490-498. [60]

Bielski, B. H. J.; Cabelli, D. E.; Arudi, R. L.; Ross, A. B. Reactivity of HO2/O2- radicals in

aqueous-solution. J. Phys. Chem. Ref. Data 14 (1985) 1041-1100.

23

[61]

Winterbourn, C. C.; Hampton, M. B.; Livesey, J. H.; Kettle, A. J. Modeling the

reactions of superoxide and myeloperoxidase in the neutrophil phagosome: Implications for microbial killing. J. Biol. Chem. 281 (2006) 39860-39869. [62]

Gebicki, J. M.; Nauser, T.; Domazou, A.; Steinmann, D.; Bounds, P. L.; Koppenol, W.

H. Reduction of protein radicals by gsh and ascorbate: Potential biological significance. Amino Acids 39 (2010) 1131-1137. [63]

Jovanovic, S. V.; Simic, M. G. Repair of tryptophan radicals by antioxidants. J. Free

Radic. Biol. Med. 1 (1985) 125-129. [64]

Thomas, A. H.; Serrano, M. P.; Rahal, V.; Vicendo, P.; Claparols, C.; Oliveros, E.;

Lorente, C. Tryptophan oxidation photosensitized by pterin. Free Radic. Biol. Med. 63 (2013) 467-475. [65]

Sormacheva, E. D.; Sherin, P. S.; Tsentalovich, Y. P. Dimerization and oxidation of

tryptophan in UV-A photolysis sensitized by kynurenic acid. Free Radic. Biol. Med. 113 (2017) 372-384. [66]

Buettner, G. R. The pecking order of free radicals and antioxidants: Lipid

peroxidation, alpha-tocopherol, and ascorbate. Arch. Biochem. Biophys. 300 (1993) 535543. [67]

DeFelippis, M. R.; Murthy, C. P.; Broitman, F.; Weinraub, D.; Faraggi, M.; Klapper, M.

H. Electrochemical properties of tyrosine phenoxy and tryptophan indolyl radicals in peptides and amino acid analogs. J. Phys. Chem. 95 (1991) 3416-3419. [68]

Kettle, A. J.; Carr, A. C.; Winterbourn, C. C. Assays using horseradish peroxidase and

phenolic substrates require superoxide dismutase for accurate determination of hydrogen peroxide production by neutrophils. Free Radic. Biol. Med. 17 (1994) 161-164.

24

[69]

Candeias, L. P.; Wardman, P.; Mason, R. P. The reaction of oxygen with radicals from

oxidation of tryptophan and indole-3-acetic acid. Biophys. Chem. 67 (1997) 229-237. [70]

Nakagawa, M.; Watanabe, H.; Kodato, S.; Okajima, H.; Hino, T.; Flippen, J. L.; Witkop,

B. A valid model for the mechanism of oxidation of tryptophan to formylkynurenine—25 years later. Proc. Natl. Acad. Sci. USA 74 (1977) 4730-4733. [71]

Nakagawa, M.; Okajima, H.; Hino, T. Photosensitized oxygenation of N-

methoxycarbonyltryptamines. A new pathway to kynurenine derivatives. J. Am. Chem. Soc. 99 (1977) 4424-4429. [72]

Josimović, L.; Janković, I.; Jovanović, S. V. Radiation induced decomposition of

tryptophan in the presence of oxygen. Radiat. Phys. Chem. 41 (1993) 835-841. [73]

Jayson, G. G.; Scholes, G.; Weiss, J. Formation of formylkynurenine by the action of

X-rays on tryptophan in aqueous solution. Biochem. J. 57 (1954) 386-390. [74]

Davies, M. J.; Fu, S. L.; Wang, H. J.; Dean, R. T. Stable markers of oxidant damage to

proteins and their application in the study of human disease. Free Radic. Biol. Med. 27 (1999) 1151-1163. [75]

Finley, E. L.; Dillon, J.; Crouch, R. K.; Schey, K. L. Identification of tryptophan oxidation

products in bovine α-crystallin. Protein Sci. 7 (1998) 2391-2397. [76]

Sherin, P. S.; Zelentsova, E. A.; Sormacheva, E. D.; Yanshole, V. V.; Duzhak, T. G.;

Tsentalovich, Y. P. Aggregation of alpha-crystallins in kynurenic acid-sensitized UVA photolysis under anaerobic conditions. Phys. Chem. Chem. Phys. 18 (2016) 8827-8839. [77]

Thomas, S. R.; Stocker, R. Redox reactions related to indoleamine 2,3-dioxygenase

and tryptophan metabolism along the kynurenine pathway. Redox Rep. 4 (1999) 199-220.

25

[78]

Rafice, S. A.; Chauhan, N.; Efimov, I.; Basran, J.; Raven, E. L. Oxidation of L-

tryptophan in biology: A comparison between tryptophan 2,3-dioxygenase and indoleamine 2,3-dioxygenase. Biochem. Soc. Trans. 37 (2009) 408-412. [79]

Morgan, P. E.; Pattison, D. I.; Davies, M. J. Quantification of hydroxyl radical-derived

oxidation products in peptides containing glycine, alanine, valine, and proline. Free Radic. Biol. Med. 52 (2012) 328-339. [80]

Morgan, P. E.; Pattison, D. I.; Hawkins, C. L.; Davies, M. J. Separation, detection, and

quantification of hydroperoxides formed at side-chain and backbone sites on amino acids, peptides, and proteins. Free Radic. Biol. Med. 45 (2008) 1279-1289. [81]

Storkey, C.; Pattison, D. I.; Gaspard, D. S.; Hagestuen, E. D.; Davies, M. J. Mechanisms

of degradation of the natural high-potency sweetener (2R,4R)-monatin in mock beverage solutions. J. Agric. Food Chem. 62 (2014) 3476-3487.

Fig. 1. Representative kinetic data for the reaction of GlyTrp● with O2●- generated during pulse radiolysis. A) Change in the absorption of GlyTrp● over time (0 - 200 µs) after an O2 saturated-solution containing GlyTrp (200 µM), NaN3 (3 mM) and formate (50 mM) in phosphate buffer (5 mM, pH 7.4) was exposed to 200 ns electron pulses with a total radiation dose of 16 Gy. The white line shows the pseudo first order kinetics fit to the experimental data. B) Plot of kobs (determined by fitting single-exponential decay functions to data similar to that shown in A)) against the initial O2●- concentration. The second–order rate constant (see Table 1) was determined from the gradient of the resulting line. C) Plot of second-order rate constants for reaction between a range of peptide-derived Trp● and O2●-, against total peptide charge, showing a positive correlation between increasing positive charge and the second order reaction rate constants. Fig. 2. Detection of organic peroxides (FOX assay reactive species) formed on irradiation of Trp-containing peptides (500 µM) to increasing doses of radiation (500 or 1000 Gy, with the 1000 Gy dose delivered at either a low, 18 Gy min-1, or high, 40 Gy min-1, irradiation dose

26

rates). Trp-containing peptides (as indicated on the panels) were irradiated in the presence of O2 and NaN3 (7.5 mM) and formate (125 mM) in 5 mM phosphate buffer (pH 7.4) at 22 o

C. Catalase (50 µL per 10 mL of sample; 1 mg mL-1) was added to each sample on cessation

of irradiation to remove any H2O2 before analysis of peroxide levels using the FOX assay. The FOX assay measurements were converted to H2O2 equivalents using a factor of 2.22 based on concentrations determined by iodometric assay as shown in Supplementary Fig. 1. * Indicates a significant difference (p < 0.05) compared to the control (non-irradiated samples). No statistical difference was detected between the 1000 Gy irradiated samples exposed at either the low or high irradiation dose rates. Error bars represent SD from n = 3 independent determinations. Fig. 3. Reaction of Trp●, from the indicated Trp-containing peptides, with O2●- results in the formation of (A, C) NFK and (B, D) Kyn. Control and irradiated (1000 Gy total dose, at a dose rate of 40 Gy min-1, at pH 7.4 and 22 oC in 5 mM phosphate buffer) samples were digested at 37 °C overnight by addition of pronase, before filtration and analysis by LC-MS (see Materials and methods). Panels A) and B) show representative selected ion chromatograms from the LC-MS analyses for NFK (m/z 237; panel A) and Kyn (m/z 209; panel B) for control (black) and irradiated (red) GlyTrp samples. Panels C) and D) show the concentrations of NFK (panel C) and Kyn (panel D) determined for the control (black bars) and irradiated samples (grey bars). Fig. 4. Reaction of Trp●, from GlyTrp, with O2●- results in the formation of multiple oxidized species as detected by LC-MS (reported as peak areas in arbitrary units). Irradiated samples (see legend to Fig. 3) were analyzed by mass spectrometry directly after the cessation of irradiation (black bars), after incubation overnight at 37 °C (white bars), or after reduction by addition of NaBH4 (grey bars). A) Yield of NFK (m/z +32 compared to the parent compound) detected from GlyTrp under the different incubation conditions. B) Yield of Kyn (m/z +4 compared to the parent compound) formed from GlyTrp under the different incubation conditions. C) Yield of species with m/z + 16 compared to the parent compound, and assigned to an alcohol, detected under the different incubation conditions. D) Yield of additional species with m/z + 16 compared to the parent compound, and assigned to a second alcohol, formed only after NaBH4 treatment of the samples.

27

Fig. 5. Proposed mechanisms of radical mediated degradation reactions of Trp-derived radicals and kinetic data for the competing pathways of the initial Trp●.

Fig. 1

28

Fig. 2

29

Fig. 3

30

Fig. 4

31

Fig. 5

Graphical Abstract

32

Table 1. Second-order rate constants (k) for range of Trp-containing peptide radicals with O2●- determined in the current study at pH 7.4 and 22 oC in 5 mM phosphate buffer, and the corresponding dimerization rate constants for the Trp● radicals generated from the Trpcontaining peptides (data from [49]), as determined by pulse radiolysis studies. Errors represent the 95 % confidence intervals from n = 3 independent experiments.

Peptide

O2●- addition

Trp● dimerization a

k / 109 M-1 s-1

k / 108 M-1 s-1

LysTrpGlyLys

2.2 ± 0.1

2.0 ± 0.1

LysTrpLys

1.9 ± 0.1

2.0 ± 0.1

cycloGlyTrp

1.6 ± 0.1

6.0 ± 0.3

GlyLysArgTrpGly

1.4 ± 0.2

2.1 ± 0.1

GlyTrpGlyGly

1.5 ± 0.1

3.9 ± 0.1

GlyTrpGly

1.5 ± 0.1

4.5 ± 0.1

LysTrp

1.1 ± 0.1

3.4 ± 0.1

NAcTrpOMe

1.1 ± 0.1

6.4 ± 0.2

ValTrp

0.9 ± 0.1

3.6 ± 0.1

MetTrp

0.9 ± 0.1

3.3 ± 0.1

ArgTrp

0.9 ± 0.1

3.6 ± 0.1

GlyTrp

0.9 ± 0.1

4.1 ± 0.1

HisTrp

0.8 ± 0.1

3.5 ± 0.2

GluTrp

0.8 ± 0.1

2.3 ± 0.1

a

Dimerization rate constants from [49].

Table 2. Trp oxidation products detected after irradiation of Trp-containing peptides in the presence of O2●- at pH 7.4 and 22 oC in 5 mM phosphate buffer in the presence of NaN3 (7.5 mM) and formate (125 mM) and subsequent forced decomposition by either heat 33

treatment or NaBH4 reduction (see Materials and methods). Materials were characterized on the basis of m/z, and fragmentation patterns after LC-MSn analysis (see Materials and methods). Increases or decreases in yield are indicated by arrows. Data are from 6 determinations from 2 independent irradiations.

GluTrp Relative peak areas m /z

Retenti

on time / min

Irrad iated

Irradi

vs ated

Irrad

and iated

non-

incubated vs NaBH4

irradiated

irradiated

Assignment

and vs

irradiated

3

14.2

↓

-

-

Parent GluTrp

3

8.6

↑

-

-

Kynurenine species

34

38

? 3

10.6

↑

↑

-

Kynurenine species

3

11

↑

-

↑

Alcohol

38

50

(degradation product of OOH) 3

12.5

↑

-

-

Alcohol

3

11.45

↑

↓

↓

NFK (loss of 46

50

66

m/z) GlyTrp Relative peak areas m

/z

Retenti

on time / min

Irrad iated non-

Irradi

vs ated

and iated

incubated vs NaBH4

34

Irrad and vs

Assignment

irradiated

irradiated

irradiated

2

14.3

↓

-

-

Parent GlyTrp

2

10.5

↑

↑

-

Kynurenine species

2

5.9

↑

-

↑

Alcohol

62

66

78

(degradation product of OOH) 2

10.1

↑

-

↑

Alcohol

78

(degradation product of OOH) 2

12.5

↑

-

↑

Alcohol

78

(degradation product of OOH) 2

7.4

-

-

↑

Alcohol

78

(degradation product of OOH) 2

8.1

-

-

↑

Alcohol

78

(degradation product of OOH) 2

11.2

↑

↓

↓

NFK (loss of 46

94

m/z) LysTrp Relative peak areas m

/z

Retenti

on time

Irrad iated

Irradi

vs ated

Irrad

and iated

non-

incubated vs NaBH4

irradiated

irradiated

Assignment

and vs

irradiated

3

13.8

↓

-

-

Parent LysTrp

3

10.8

↑

↑

-

Kynurenine species

33

35

37 3

6.4

↑

-

↑

Alcohol

49

(degradation product of OOH) 3

8.8

↑

-

↑

Alcohol

49

(degradation product of OOH) 3

11.4

↑

↓

↓

Inconclusive;

65

loss

of m/z 111 6.9

↑

↓

-

Diol

Irrad

Assignment

LysTrpLys Relative peak areas m /z

Retenti

on time

Irrad iated

Irradi

vs ated

and iated

non-

incubated vs NaBH4

irradiated

irradiated

and vs

irradiated

4

10.3

↓

-

-

Parent LysTrpLys

4

7.4

↑

↑

↓

Kynurenine species

4

8.7

↑

↑

↓

Kynurenine species

61

65

65

? 4

6.5

↑

-

↑

77

Alcohol (degradation product of OOH)

4

6.8

↑

↓

↓

93

Inconclusive;

loss

of m/z 18 4

9.6

↑

↓

93

↓

Inconclusive; of m/z 110

ValTrp Relative peak areas

36

loss

m /z

Retenti

on time

Irrad iated

Irradi

vs ated

Irrad

and iated

non-

incubated vs NaBH4

irradiated

irradiated

Assignment

and vs

irradiated

3

15.5

↓

-

-

Parent ValTrp

3

10.7

↑

-

-

Kynurenine species

04

08

? 3

12.8

↑

↑

-

Kynurenine species

3

9.2

↑

-

↑

Alcohol

08

20

(degradation product of OOH) 3

10.6

↑

-

↑

20

Alcohol (degradation product of OOH)

3

12.6

↑

-

↑

20

Alcohol (degradation product of OOH)

3

14.4

↑

-

-

Alcohol ?

3

13.25

↑

↓

↓

NFK (m-z of -46)

20

36

Table 3. Reactions of tryptophan with superoxide produce a near stoichiometric conversion of Trp● to peroxide species as detected by FOX assay, with approximately 10 % conversion of Trp● to NFK.

Initial radical

RO OH / µM

N FK / µM 37

Kyn / µM

concentration µM

/

O 2

●-

392 Tr

p●

168 Gl

uTrp

166 ± 50

Gl yTrp

157 ± 40

Ly sTrp

1 7±1

171 ± 46

Ly sTrpLys V alTrp Gl yTrpGly

1 7±1

4 ±1

194 ± 49

1 7±3

211 ± 100

4 ± 0.4

186 ± 65

1 8±2

0.2 2 ± 0.04 0.2 7 ± 0.04 0.1 5 ± 0.1 0.3 8 ± 0.08 0.2 3 ± 0.1 0.4 3 ± 0.05

HIGHLIGHTS     

Tryptophan (Trp) residues are a major sink for oxidants on peptides and proteins Trp radicals in peptides undergo fast reactions with O2●- (k ~ 109 M-1s-1) Trp hydroperoxides are major products of Trp● + O2●- reactions in peptides Hydroperoxide degradation gives N-formylkynurenine, kynurenine and alcohols Hydroperoxide formation competes favourably with dimerization and O2 addition

38