Efficient depletion of ascorbate by amino acid and protein radicals under oxidative stress

Free Radical Biology and Medicine 53 (2012) 1565–1573

Contents lists available at SciVerse ScienceDirect

Free Radical Biology and Medicine journal homepage: www.elsevier.com/locate/freeradbiomed

Original Contributions

Efficient depletion of ascorbate by amino acid and protein radicals under oxidative stress Anastasia S. Domazou a, Viviane Zelenay a,b, Willem H. Koppenol a, Janusz M. Gebicki b,n a b

Institute of Inorganic Chemistry, Department of Chemistry and Applied Biosciences, Swiss Federal Institute of Technology, Zurich CH-8093, Switzerland Department of Biological Sciences, Macquarie University, Sydney 2109, Australia

a r t i c l e i n f o

abstract

Article history: Received 15 March 2012 Received in revised form 27 June 2012 Accepted 3 August 2012 Available online 11 August 2012

Ascorbate levels decrease in organisms subjected to oxidative stress, but the responsible reactions have not been identified. Our earlier studies have shown that protein C-centered radicals react rapidly with ascorbate. In aerobes, these radicals can react with oxygen to form peroxyl radicals. To estimate the relative probabilities of the reactions of ascorbate with protein C- and O-centered radicals, we measured by pulse radiolysis the rate constants of the reactions of C-centered radicals in Gly, Ala, and Pro with O2 and of the resultant peroxyl radicals with ascorbate. Calculations based on the concentrations of ascorbate and oxygen in human tissues show that the relative probabilities of reactions of the C-centered amino acid radicals with O2 and ascorbate vary between 1:2.6 for the pituitary gland and 1:0.02 for plasma, with intermediate ratios for other tissues. The high frequency of occurrence of Gly, Ala, and Pro in proteins and the similar reaction rate constants of their C-centered radicals with O2 and their peroxo-radicals with ascorbate suggest that our results are also valid for proteins. Thus, the formation of protein C- or O-centered radicals in vivo can account for the loss of ascorbate in organisms under oxidative stress. & 2012 Elsevier Inc. All rights reserved.

Keywords: Amino acid radicals Protein radicals Protein oxidation Protein repair Ascorbate Ascorbate radicals Antioxidant Rate constant Pulse radiolysis Free radicals

Introduction Aerobic organisms are continuously exposed to partially reduced oxygen species, or PROS. This term provides a more precise definition of the species involved than the commonly used ROS (reactive oxygen species), because it includes only the intermediates formed in stepwise reduction of oxygen: O2  , H2O2, and the hydroxyl radical (HO). When their formation exceeds the organism’s antioxidant capacity, the organism becomes subject to oxidative stress, which may lead to activation of specific signaling pathways and temporary or permanent damage to vital cell components, resulting in disease or death (reviewed in [1]). The link between oxidative stress and pathology is now firmly established and is supported by the identification of over 50 diseases and debilitating conditions it can cause or aggravate. However, although many of the pathological outcomes of the oxidative stress are known, control of the damaging actions of PROS in vivo remains elusive, largely because the identities of the molecular pathways linking oxidative stress and pathology are as yet not established. In particular, the nature of the principal initial biological targets of the

n

Corresponding author. Fax: þ61 2 9850 8245. E-mail address: [email protected] (J.M. Gebicki).

0891-5849/$ - see front matter & 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.freeradbiomed.2012.08.005

PROS is still unknown. Such targets must either be necessary for cell functioning or have the capacity to pass the PROS-induced damage to vital molecules. It is clear that a major component of cells and tissues that must be preserved or replenished to allow the organism to function when challenged by oxidative stresses is made up of their antioxidant defenses. These defenses are principally enzymatic or made up of the ubiquitous endogenous ascorbate (Hasc  ; systematic IUPAC name monohydrogen ascorbate) and GSH (glutathione). There is considerable evidence demonstrating the depletion of both these antioxidants in living organisms subjected to oxidative stress [1]. Loss of Hasc  may be especially significant because of its effectiveness as an antioxidant with a low reduction potential and the need to be replaced by diet in primates lacking an essential enzyme in its synthetic pathway. Examples of evidence of the loss of Hasc  under oxidative stress include results obtained with cells and tissues [2–4] and in in vivo studies of rat ischemic or damaged brains [5,6]. The depletion of Hasc  has also been reported in humans with severe asthma, advanced renal and bowel disease, and diseases typical of old age such as vascular dementia, Alzheimer disease, and Parkinson disease [7–10]. However, the molecular mechanisms responsible for the depletion of Hasc  have not been identified. In recent years, we have been investigating the theory that proteins (PrH) oxidized by PROS can act as initiators of further

1566

A.S. Domazou et al. / Free Radical Biology and Medicine 53 (2012) 1565–1573

damage in cells. Proteins are clearly major potential targets of PROS, because they constitute about 70% of the organic mass of cells, plasma, and other tissues, and they react readily with many physiologically important oxidants [11]. Their initial reaction with many PROS results in the formation of protein radicals (Pr) located on the amino acid residues (AAH) (Reaction (1)), which in the presence of oxygen can be rapidly converted to peroxyl radicals (PrOO; Reaction (2)). Subsequent reduction leads to the generation of semistable protein hydroperoxides (PrOOH) in yields exceeding 50% of the attacking HO or peroxyl radical (ROO) [11,12]: PrHþ R 2Pr þ RH

ð1Þ

Pr þO2-PrOO 

PrOO þRH-PrOOHþR

(2) 

(3)

Reaction (1), oxidation by a radical R, involves mainly C-centered radicals. The reaction sequence (1)–(3) was shown to occur in a range of stressed biological systems [13,14], in which the PrOOH or its decomposition products have the potential to impair the normal functions of DNA, lipids, enzymes, and antioxidants [11]. In cells and tissues, the reduction of both the Pr and the PrOO is highly likely, because the endogenous antioxidants are normally fairly concentrated and Reactions ( 1) and (3) are fast. Earlier studies have measured the rate constants of the reduction of the aromatic amino acid and protein radicals in the absence of O2 by low-molecular-weight compounds such as urate, ascorbate, GSH, flavonoids, and vitamin E and its analogue, Trolox (reviewed in [15]). It is important to note that the reduction of O-centered amino acid radicals generates reactive species capable of propagating the initial PROS-induced damage. In contrast, reaction of the C-centered radicals with agents such as Hasc  represents a repair, generating a stable product. The biological consequences of the reduction of the protein C- or O-centered radicals can be severe, if they result in the loss of the most important endogenous antioxidants, Hasc  and GSH. This can lead to the lowering of the organism’s antioxidant potential, rendering it increasingly susceptible to further damage. If protein radicals play a role in this loss, the assessment of the relative probabilities of the occurrence of reaction (  1) or (3) with Hasc  as RH requires the knowledge of its concentrations and location in vivo and of the rate constants of its reactions with protein radicals. The distribution and concentrations of Hasc  in human tissues are well known, but data on the rate constants is limited to reactions with selected AAH residues. Early studies using fast kinetic techniques involved the reactions of Trp and Tyr radicals (Trp and TyrO, respectively) generated in lysozyme by pulse radiolysis, which oxidized Hasc  with k¼ 1.1  107 and 8  107 M  1 s  1, respectively [16,17]. We recently extended these measurements to the reactions of Hasc  with Trp and TyrO located in insulin, pepsin, chymotrypsin, b-lactoglobulin, and bovine serum albumin. The k values for protein–Trp reactions were (2.2–18)  107 M  1 s  1 and for protein–TyrO (4–290)  105 M  1 s  1, giving a reasonable average for AA þHasc  of  6  107 M  1 s  1 [18]. These results show that oxidation of Hasc  by radicals generated in proteins have the potential to account for much of its loss under oxidative stress [3–6,8–10]. However, the measurements of the k values had to be confined to reactions of the Trp and Tyr residues, because only their radicals have readily accessible absorption spectra. In addition, because Trp and TyrO do not react with O2 or react only slowly [19,20], the potentially important reactions of PrOO generated in Reaction (2) could not be measured. These limitations can be significant for cells, in which radicals are likely to be formed also in AAH other than Trp and Tyr and in which peroxyl radicals may be the

most important agents of molecular damage [21]. We have now investigated the kinetics of the formation and decay of the peroxyl radicals (AAOO) of Gly, Ala, and Pro derivatives and of their reactions with Hasc  . The process leads to loss of Hasc  with formation of the ascorbyl radical, asc  , and generates amino acid (protein) hydroperoxides (AAOOH), with the potential to propagate the damage to other biological targets: AAOO þHasc  -AAOOHþ asc 

(4)

When extended to living organisms, the overall results suggest that Hasc  can be efficiently oxidized by either C- or O-centered protein radicals, with the prevailing reaction depending mainly on the local concentration of the vitamin. We also report measurements of the rate constants of reactions of the peroxyl radicals of the Gly, Ala, and Pro derivatives with tetramethyl phenylene diamine (TMPD). On oxidation, TMPD converts to a fairly stable radical, TMPD þ (Wurster’s blue), with strong absorbance at 565 nm [22]. Our results can be used to measure by competitive kinetics the rate constants of reactions of these AAOO with compounds that cannot be studied directly by fast kinetic techniques. This should allow the evaluation of relative biological effectiveness of a wide variety of potential biological antioxidants.

Experimental methods Materials All chemicals were of the highest quality available and were used without further purification. Water was purified in a Millipore Milli-Q unit. Sodium phosphate and sodium azide were supplied by Fluka (Buchs, Switzerland). The ascorbic acid and amino acids and their N-acetyl amides were from Fluka or Bachem (Bubendorf, Switzerland). Pulse radiolysis Electron pulses of 2.0 MeV energy and 50 ns duration were generated by a Febetron 705 accelerator (Titan Systems Corp., San Leandro, CA, USA). The light source was a 75-W xenon arc ¨ lamp (Hamamatsu, Schupfen, Switzerland) and the optical pathlength of the quartz cell containing the solution to be irradiated (Hellma GmbH & Co KG, Zumikon, Switzerland) was 1 cm. The cell volume was 300 ml, and in most experiments a fresh sample was introduced before each pulse. The detection system consisted of a Roper Scientific Acton SP300i monochromator (Ottobrunn, Germany) and a Hamamatsu R928 photomultiplier, a Femto Messtechnik GmbH DHPCA-200 or HCA-100M-50k-C amplifier (Berlin, Germany), and a Yokogawa Electric Corp. DL7100 digital storage oscilloscope (Tokyo, Japan). The energy dose per pulse was 4–150 Gy, as determined by thiocyanate dosimetry [23]. Generation of amino acid radicals Radiolysis of dilute aqueous solutions generates oxidizing  (HO) and reducing (eaq þ H) radicals in similar yields [24]. In solutions saturated with 24 mM N2O (IUPAC systematic name  oxidodinitrogen), the hydrated electrons (eaq ) are converted to  9 1 1 HO with rate constant k5 ¼ 9.1  10 M s [24]. The concentration of HO generated in such solutions at neutral pH, is, then, close to 0.55 mM per Gy absorbed (1 Gy¼1 J kg  1):  þN2OþH2O-HO þN2 þ OH  eaq

(5)

The primary, C-centered, amino acid radicals (AA) were generated by irradiating AAH solutions, saturated by N2O; in

A.S. Domazou et al. / Free Radical Biology and Medicine 53 (2012) 1565–1573

Kinetics of formation and reactions of amino acid radicals N-acetyl-amino acid solutions were studied at pH 2 and pH 7.3, and the solutions of the N-acetylamides were buffered with 5 mM phosphate at pH 7.3. Formation, decay, and reactions of radicals were followed by measurement of absorbance changes following the pulse. The smooth lines drawn in all figures that show absorbance changes after the electron pulse were generated by analysis of the results with Kaleidagraph from Synergy Software (Reading, PA, USA). Error bars in the figures represent standard deviations from the mean value, and the second-order rate constants derived from linear plots of kobs versus reactant concentration are given as mean value7tss for 95% probability. The pulsed solutions needed to contain concentrations of AAH high enough to ensure scavenging of all the HO and avoid direct reaction of HO with Hasc  or TMPD. The reaction Hasc  þHO-asc  þH2O

(7)

as well as that of HO with TMPD both have a rate constant of 1010 M  1 s  1 [22], so with up to 370 mM Hasc  or TMPD and the AAH concentration of 50 mM, the AAH scavenged 490% of the HO. In the experiments performed in the absence of O2, the radiation dose was usually 10–70 Gy. The dose used to generate the AA spectra was  60 Gy for Gly and Pro derivatives and 10–20 Gy for Ala. Low doses were used to study the reaction of AA with O2 (10–30 Gy) and those of AAOO with Hasc  (3–15 Gy; at low Hasc  concentration o5 Gy) and with TMPD (4–30 Gy), to avoid radical–radical recombination and ensure pseudo-first-order conditions. The reactions of AAOO with Hasc  or TMPD were studied in solutions degassed with a 1:4 mixture of O2 and N2O.

Results Our principal experimental objectives were to test whether amino acid oxygen-centered radicals react with Hasc  and to measure the kinetics of these reactions. The high concentrations of AAH needed to ensure scavenging of all the HO imposed a limit on the choice of AAH to be studied. We selected the N-acetylamides of Ala (N-Ac-Ala-NH2), Gly (N-Ac-Gly-NH2), and Pro (N-Ac-Pro-NH2), because of their high solubilities, the absence of additional functional groups, the high frequency of occurrence in proteins, and their having the chemical form of protein

6 3

At neutral pH, the hydrogen atoms (H) do not alter the overall yield of AA.

The amino acid C-centered radicals were generated in solutions saturated with N2O, in which their absorption spectra can be observed in the range of 240–500 nm. Fig. 1A shows the spectra of the radical N-Ac-Ala recorded at pH 2 and of N-Ac-Ala-NH2 at pH 7.3. The spectra showed no dose dependence within the range 4–60 Gy. The spectrum at acidic pH corresponds closely to that recorded in an earlier pulse radiolytic study of (Gly6) at pH 1.7 [27], whereas that recorded at pH 7.3 shows a second peak at 330 nm, reflecting the influence of the charged carboxyl group. Based on the assumption of complete scavenging of the HO by the AAH and formation of a-C radicals only, the value of the molar absorption coefficient (e) for N-Ac-Ala-NH2 at 260 nm was  9000 M  1 cm  1, close to that recorded in an earlier study [27]. Fig. 1B shows the kinetics of the decay of N-Ac-Ala-NH2 at 350 nm. The decay—as well as the decay of the N-AcAla—followed second-order kinetics, demonstrating that it was due to radical–radical recombination. The rate constants of the decay were not dependent on radiation dose and are listed in Table 1. The spectrum of the N-Ac-Gly-NH2 was similar to that of the corresponding N-Ac-Ala-NH2. The N-Ac-Gly-NH2 also

Absorbance (×10 )

(6)

Formation and decay of amino acid radicals in the absence of O2

4

a

2 b 0 200

300 400 Wavelength (nm)

500

500 1000 Time (µs)

1500

120 3

AAHþHO - AA þH2O

residues [26]. In each case, we generated the spectrum of the C-centered amino acid radical formed in the electron pulse and measured the rate constant of its disappearance in the absence of O2, its reaction with O2, and the rate constant of the reaction of the resultant O-centered radicals, AAOO, with Hasc  and with TMPD. Experiments were also carried out with the N-acetyl derivatives of Ala (N-Ac-Ala) and Pro (N-Ac-Pro) having underivatized carboxyl groups.

Absorbance (×10 )

experiments involving AAOO, mixtures of O2 and N2O between 1:14 and 1:4 were employed. Gas saturation was achieved by repeated evacuation of the solutions followed by equilibration with the appropriate gas mixture. The solution was then taken up into a gas-tight syringe and transferred to the quartz cell located in the path of the high-energy electrons. Instead of free AAH, we used the corresponding N-acetylamide and, in some cases, the N-acetyl derivatives as better models of these residues in proteins and to avoid the influence of the charged groups. The AA were generated by H abstraction and their yields equaled the yields of HO, provided the parent AAH concentration was sufficient to ensure complete scavenging of the HO in Reaction (6) and there were no other solutes able to compete:

1567

80 40 0

0

Fig. 1. (A) Absorbance spectrum of N-Ac-Ala-NH2 at pH 7.3 (curve a) and N-AcAla at pH 2 (curve b). The amino acid concentration was 50 mM and the 5 mM phosphate solutions at pH 7.3 were saturated with N2O. The spectra were recorded 2 ms after the pulse and the results normalized for energy dose of 1 Gy. (B) The bimolecular decay kinetics of the N-Ac-Ala-NH2 generated by 19.2 Gy (10.5 mM HO). The decay was monitored at 350 nm. The red line is the fit to the second-order rate law. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

1568

A.S. Domazou et al. / Free Radical Biology and Medicine 53 (2012) 1565–1573

disappeared by radical–radical recombination, with the rate constant shown in Table 1. In contrast, the absorbance of the radical of N-Ac-Pro-NH2 (N-Ac-Pro-NH2) showed slightly different characteristics. Compared to the N-Ac-Ala-NH2, at pH 7.3, the prominent peak was at 250, not 260 nm; the lower peak was shifted from 330 to 360 nm (Fig. 2A); and a new peak appeared at 290 nm. Also, unlike the N-Ac-Ala-NH2, N-Ac-Ala, and N-Ac-Gly-NH2, which decayed completely in less than 2 ms with no remaining absorbance, significant absorbance remained after the complete decay of the N-Ac-Pro-NH2 in a radical–radical recombination. Fig. 2B shows the second-order decay kinetics at the 360- and 250-nm absorbance peaks, which demonstrate a persistent 250-nm absorbance due to the formation of a stable product. Assuming complete scavenging of the HO by the AAH and formation of only one radical species, the e values for N-Ac-Pro-NH2 could be calculated, giving maxima of 1700 M  1 cm  1 at 360 nm and 2500 M  1 cm  1 at 250 nm. The rate constant of the second-order decay is shown in Table 1. The

Table 1 Rate constants of amino acid radical recombination and of their reactions with O2. Amino acid (AAH)

2k (M  1 s  1) (2 AA - products)

k (M  1 s  1) (AA þO2-AAOO)

N-Ac-Gly-NH2 N-Ac-Ala-NH2 N-Ac-Pro-NH2 N-Ac-Ala N-Ac-Ala N-Ac-Pro

(8.7 70.6)  108 (1.3 70.3)  109 (1.1 70.3)  109 (3.7 70.7)  108 (8.7 70.9)  108 (9.1 70.5)  108

(2.070.9)  109 (5.57 0.7)  109 (6.37 0.5)  108 (1.57 0.2)  109 (3.37 0.3)  109 (2.57 0.4)  108

a a

a a

Measured at pH 7.3. a

Measured at pH 2.

radical N-Ac-Pro showed similar behavior at pH 2 and decayed at a comparable rate (Table 1). Formation and decay of amino acid radicals in the presence of O2 The presence of O2 led to significant acceleration in the spontaneous decay of the AA. For 10 mM N-Ac-Ala-NH2, the half-time of its decay in the radical–radical reaction (t1/2  100 ms) decreased by a factor of  120 in the presence of 156 mM O2 (Fig. 3A). Measurements in the presence of increasing concentrations of O2 gave a series of pseudo-first-order rate constants, from which the bimolecular rate constant of Reaction (8) was derived (Fig. 3B, Table 1): N-Ac-Ala-NH2 þO2-N-Ac-AlaOO-NH2

It is important to note that, even when the rate constants of the radical–radical reactions and of the reactions of the radicals with other solutes were similar (e.g., Table 1), the generally low radical and much higher solute concentrations lead to much faster radical–solute reactions. This is clearly illustrated in the example shown in Fig. 3A, in which there was no detectable decay of the N-Ac-Ala-NH2 in the 5 ms after the pulse in the absence of O2, whereas an identical amount of the radicals decayed virtually completely within that time in reaction with the O2. In similar experiments, the N-Ac-GlyOO-NH2 was generated in solutions of the parent AAH equilibrated with mixtures of O2 and N2O. The presence of O2 accelerated the decay of the N-AcGly-NH2 formed initially, in a concentration-dependent manner, with the decay now obeying first-order kinetics. The second-order rate constant of the reaction analogous to Reaction (8) was derived from the experimental pseudo-first-order constants

120 Absorbance (×103)

Absorbance (×103)

1.5 a

1.0 0.5

b 0.0 200

a

40

500

b 0

5

10 15 20 25 30 Time (µs)

16 -1

(s )

80

×10

-5

a

obs

40

k

Absorbance (×103)

80

0

300 400 Wavelength (nm)

120

b 0

(8)

12 8 4 0

0

500 1000 Time (µs) 

1500

Fig. 2. (A) The spectrum of the N-Ac-Pro -NH2 recorded 2 ms (curve a) and 1.2 ms (curve b) after the pulse. The parent N-Ac-Pro-NH2 concentration was 50 mM, the 5 mM phosphate solutions at pH 7.3 were saturated with N2O, the energy dose was  60 Gy, and the spectra were normalized to 1 Gy. Similar results were recorded with doses between 10 and 120 Gy. (B) The decay kinetics of the N-AcPro-NH2 recorded at 250 nm (curve a) and at 360 nm (curve b). The energy doses were 64.6 and 62.3 Gy, respectively. The red line is the fit to the second-order rate law. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

0

100 200 [O ] (µM)

300

2

Fig. 3. (A) The decay kinetics of the N-Ac-Ala-NH2 in the absence (curve a) and in the presence (curve b) of 156 mM O2. The parent amino acid concentration was 50 mM, the 5 mM phosphate solutions at pH 7.3 were saturated with N2O (curve a) or 8:1 N2O:O2 (curve b), and the energy dose was 20 Gy. The red line is the single exponential fit. (B) Plot of the pseudo-first-order decay rate constants of the N-Ac-Ala-NH2 at various concentrations of O2. The second-order rate constant of the N-Ac-Ala-NH2 reaction with O2 was derived from this result. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

A.S. Domazou et al. / Free Radical Biology and Medicine 53 (2012) 1565–1573

15 3

Absorbance (×10 )

Absorbance (×10 3)

0.4

0.2

0 230

250 270 Wavelength (nm)

5

0

290

50

100 150 Time (µs)

200

300 350 400 Wavelength (nm)

450

1.6

20

Absorbance (× 10 3)

3

10

0

25 Absorbance (×10 )

1569

a

15 10 b

5

1000 2000 Time (µs)

3000

0.4

12

(Table 1). Fig. 4A shows the absorbance after the complete decay of the N-Ac-Gly-NH2 in solutions saturated with a mixture of 1:4 O2:N2O, which we attribute to N-Ac-GlyOO-NH2. Assuming 100% conversion of N-Ac-Gly-NH2 to N-Ac-GlyOO-NH2, the e value for N-Ac-GlyOO-NH2 at 270 nm was 11507150 M  1 cm  1. In the dose range used, the decay of N-Ac-GlyOO-NH2 followed second-order kinetics (Fig. 4B), with a rate constant of 2k 4  108 M  1 s  1. Measurements of the kinetics of the reaction of the N-Ac-ProNH2 with O2 gave similar results. In the presence of 190 mM O2, the half-life of the decay of 15 mM N-Ac-Pro-NH2 at 360 nm decreased by a factor of  20, indicating the rapidity of Reaction (9). Moreover, the formation of the stable product that absorbs at 250 nm was totally inhibited. The second-order rate constant k9 (Table 1) was derived as usual from the measured pseudo-first-order rate constants: (9)

N-Ac-Ala and N-Ac-Pro also reacted with O2 with slightly lower rate constants (Table 1). Reaction of amino acid peroxyl radical with ascorbate and TMPD The formation of N-Ac-AlaOO-NH2 and its reaction with Hasc  could be followed simultaneously at 360 nm because of the difference in their kinetics and the absorbance of the species involved. In solutions saturated with 1:4 mixture of O2:N2O containing Hasc  , the rapid absorbance decrease at 360 nm due to loss of the N-Ac-Ala-NH2 and formation of the practically nonabsorbing N-Ac-AlaOO-NH2 was followed by a slower absorbance increase due to the appearance of the asc  (Fig. 5A). The identity of asc  was confirmed by its characteristic spectrum [28] (Fig. 5B):

8 4

k

obs

×10

-3

-1

(s )

Fig. 4. (A) The spectrum of the N-Ac-GlyOO-NH2 after an  30-Gy energy pulse. The solution at pH 7.3 was saturated with 4:1 N2O:O2 and the absorbances are shown at 1 ms (’). (B) The decay kinetics of the N-Ac-GlyOO-NH2 (curve b, 270 nm) and N-AcProOO-NH2 (curve a, 250 nm) in solutions saturated with 4:1 N2O:O2. The amino acid concentration was 50 mM, the 5 mM phosphate solutions were at pH 7.3, and the energy doses were  14 and  20 Gy, respectively. The red line shows the fit to the second-order rate law. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

N-Ac-Pro-NH2 þO2 -N-Ac-ProOO-NH2

0.8

0.0 250

0 0

1.2

0 0

200 400 [Ascorbate] (µM)

600

Fig. 5. (A) The kinetics of the formation of the N-Ac-AlaOO-NH2 and asc  recorded at 360 nm. The pH of the 5 mM phosphate solution was 7.3, saturated with 4:1 N2O:O2, and the energy dose was  15 Gy. The concentration of N-Ac-Ala-NH2 was 50 mM and that of Hasc  230 mM. The initial fast absorbance decay shows formation of the N-AcAlaOO-NH2 and the slower increase corresponds to the formation of the asc  . The red line is the two single exponentials fit. (B) Absorbance spectrum of asc  normalized to 1 Gy after completion of reaction with the N-Ac-AlaOO-NH2. The concentration of Hasc  was 2 mM and the energy dose was  20 Gy. (C) The dependence of the pseudo-first-order rate constants of asc  formation on the concentration of Hasc  . (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Table 2 Rate constants of reactions of amino acid peroxyl radicals with Hasc  and TMPD. Radical

N-Ac-GlyOO-NH2 N-Ac-AlaOO-NH2 N-Ac-ProOO-NH2 N-Ac-AlaOO

k (M  1 s  1) þ Hasc 

þ TMPD

(2.5 70.5)  107 (8.7 70.9)  106 (1.5 70.2)  107 NM

(27 1)  108 (3.97 0.4)  107  6.5  107 (5.97 0.1)  107

Measurements carried out at pH 7.3. NM, not measured.

N-Ac-AlaOO-NH2 þHasc  -N-Ac-AlaOOH-NH2 þasc 

(10)

The value of k10 was derived from measurements of the dependence of the pseudo-first-order constants of Reaction (10) on the concentration of the Hasc  (Fig. 5C). The N-Ac-GlyOO-NH2 also oxidized the Hasc  . In solutions of N-Ac-Gly-NH2 saturated with a 1:4 mixture of O2:N2O, containing 266 mM Hasc  , formation of asc  was completed in about

1570

A.S. Domazou et al. / Free Radical Biology and Medicine 53 (2012) 1565–1573

400 ms (results not shown). The kinetics of the first-order decay of Hasc  at 285 nm and the formation of asc  at 360 nm were identical for the same Hasc  concentrations. The second-order rate constant of Reaction (11) was derived from these results (Table 2): N-Ac-GlyOO-NH2 þHasc  -N-Ac-GlyOOH-NH2 þasc 

(11)

Similar results were obtained with the N-Ac-ProOO-NH2. Measurement of the formation of the asc  at 360 nm in reaction analogous to Reaction (11) gave the rate constant listed in Table 2. The maximum yield of the asc  in reactions with the three amino acid peroxyl radicals tested was 100%. Reactions of the peroxyl radicals of the N-acetylamides of Ala, Gly, and Pro and of N-Ac-Ala with TMPD were studied in AAH solutions saturated with a 1:4 mixture of O2:N2O and containing TMPD. The fast formation of the N-Ac-AlaOO-NH2, detected by the disappearance of N-Ac-Ala-NH2 in Reaction (8) (Fig. 6A), was followed by the formation of the TMPD þ at 565 nm (Fig. 6B). This is explained by the following reaction: N-Ac-AlaOO-NH2 þTMPD þH þ -N-Ac-AlaOOHNH2 þ TMPD þ

(12)

3

Absorbance (×10 )

15 10 5 0

0

2

0

500

4 6 Time (µs)

8

3

Absorbance (×10 )

80 60 40 20 0

1000 1500 2000 Time (µs)

10 5

k

obs

×10

-3

-1

(s )

15

0

0

100 200 [TMPD] (µM)

300

Fig. 6. The reaction of N-Ac-AlaOO-NH2 with TMPD. The solutions at pH 7.3 contained 50 mM N-Ac-Ala-NH2, 60 mM TMPD and were saturated with 4:1 N2O:O2. The single exponential fits are shown (red lines). (A) The decay of the N-Ac-Ala-NH2 in reaction with O2. (B) Formation of the TMPD þ from the reaction of the TMPD with N-Ac-AlaOO-NH2 measured at 565 nm. (C) The dependence of the pseudo-first-order rate constants of TMPD þ formation on the concentration of TMPD. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

The yield of the colored radicals at high TMPD concentrations, calculated with the molar absorption coefficient of 12,500 M  1 cm  1 [29], reached 90–100% of the generated HO. The formation of TMPD þ followed first-order kinetics with the rate constant linearly dependent on TMPD concentration (Fig. 6C). The derived second-order rate constant k12 is listed in Table 2, together with the rate constants of the corresponding reactions of the N-AcGlyOO-NH2, N-Ac-ProOO-NH2, and N-Ac-AlaOO.

Discussion Formation and decay of amino acid radicals The HO oxidized the N-acetyl derivatives of Gly, Ala, Pro, and their amides by hydrogen abstraction. Rate constants of the reactions of HO with the free AAH and some of their derivatives are known [25] and the sites of location of the unpaired electrons have been determined in the cases of Gly, Ala, Val, Leu, and their N-acetyl derivatives and peptides [30,44]. In broad terms, the favored site of radical location is the a-C (C2) in Gly, with 22% of the radicals located on the acetyl group in the N-Ac-Gly. In Ala, all of the unpaired electrons locate on the b-C [30,44], with roughly equal amounts on the a-C and b-C in N-Ac-Ala and with less than 10% on the acetyl group [30]. However, in that study [30], the AA were generated not by pulse radiolysis but by the use of Ti(III)/ H2O2; it has been argued that not HO but a titanium complex oxidizes AAH to AA [45–47]. Thus, the sites of attack in the latter case—and the radicals formed—are not necessarily the same as those produced by the HO. Radical location on the a-C and d-C in oxidized Pro was reported, with the d-C believed to be the favored site in the N-derivatives of Pro [31,48,49]. The possibility of the formation of more than one AA suggests that their decay could follow mixed-order kinetics, but in this study radical disappearance followed strictly second-order kinetics (Table 1). This demonstrates the recombination either of a single kind of radical or of more than one reacting at rates that could not be experimentally separated. The results in Table 1 also show that the amide group had little effect on the rate constants of the radical recombination. With the exception of the N-Ac-Ala at neutral pH, the recombination rate constants were remarkably similar at about 109 M  1 s  1. The radicals generated on all Gly and Ala derivatives absorb below 500 nm with maxima in the 260–265 and 320–330 nm ranges, as observed earlier [27]. Based on the similarities of the shape and the intensities of the spectra that we obtained for N-Ac-Gly-NH2 and N-Ac-Ala-NH2, we assign the location of the unpaired electron entirely on the a-C in N-Ac-Ala-NH2 and N-AcAla, in agreement with previous reports [27]. If our assignment is not correct, and radicals are also located on the b-C, the e values for N-Ac-Ala-NH2 and N-Ac-Ala and the k values for the radical– radical decay we report here should be considered as lower limits for the correct values. Indeed, radicals located on b- or g-C relative to a carboxylic or carbonyl group absorb below 240 nm [50–52]. Interestingly, although the spectra of N-Ac-Pro-NH2 and N-Ac-Pro are similar in shape to those of the radicals of Gly and Ala derivatives, they are much less intense, with maxima shifted to 250 and 360 nm (and a third peak at 290 nm). Such differences have been observed for the a-C radicals in the simple tripeptides of Gly, Ala, and sarcosine (Sac) [53,54]: The first two show maxima at 260–265 and 330 nm, whereas (Sac)3 absorbs principally at 245 and 355 nm. Shifts similar to those for the 330-nm band have been also observed in amides for radicals located on the C atoms adjacent to N, and they are related to the degree of alkylation of N. These radicals show absorbance maxima at 238– 247 and 340–390 nm. The more alkylated the N, the more the

A.S. Domazou et al. / Free Radical Biology and Medicine 53 (2012) 1565–1573

340-nm band is shifted to longer wavelengths [51,52]. In N-acetyl derivatives of Pro, the a-C radicals are expected to have spectra similar to that of (Sac)3 and the d-C radicals similar to that of a-C (relative to N) radicals in N,N-dialkyl amides. Because both types of radicals show maxima at the same l regions, it cannot be concluded solely from the spectroscopic data whether only one (and which) or both of them are formed. Formation of d-C radicals seems more probable for the protected Pro derivatives, because of severe steric constraints associated with the planarity of the a-C radicals [31,48,49]. However, formation of b- and g-C radicals absorbing below 240 nm is also possible. The e and the k values for the radical–radical recombination for N-Ac-Pro-NH2 and N-Ac-Pro were derived under the assumption of the formation of a single radical type, a- or d-C. If more than one type of radical is formed, they represent only lower limits of the real values. In O2-free solutions, the absorbance decayed to baseline in the cases of N-Ac-Ala-NH2, N-Ac-Ala, and N-Ac-Gly-NH2, but N-AcPro-NH2 and N-Ac-Pro gave stable absorbance after  1 ms following the pulse (Fig. 2B). For Gly and Ala, only a small fraction of the initial C-centered radicals form dimers, whereas the rest disappear in radical–radical reactions resulting in deamination and formation of a range of low-molecular-weight fragments [32]. In contrast, N-Ac-Gly and N-Ac-Ala disappear without N–C bond cleavage [32]. The identity of the stable product of the secondorder decay of N-Ac-Pro-NH2 and N-Ac-Pro has not been reported to our knowledge. Whereas previous studies identified a range of products of the oxidation of Pro by HO, the experiments were carried out in the presence of O2, which was involved in the generation of the final products [33–36]. In the presence of O2, the initial products of its reaction with C-centered radicals are AAOO. The course of their formation (Reactions (8) and (9)) was readily followed because the AAOO had no significant absorbance compared to the parent radicals in the 230–450 nm region (Fig. 1A and 3A). Reaction rate constants of all the parent AA with O2 were close to the generally accepted value of  2  109 M  1 s  1 for C-centered radicals (Table 1) [24]. The AAOO are unstable, but their reactions with solutes such as Hasc  normally predominate over radical–radical or unimolecular decomposition at high solute and relative low radical concentration [24]. Reactions of amino acid peroxyl radicals with ascorbate and TMPD The radiation doses and the concentrations of AAH, O2, and Hasc  were arranged to ensure that virtually all the HO were scavenged by the AAH, the resultant AA reacted with O2, and the

1571

AAOO produced were scavenged by the Hasc  . The large differences between the rate constants of the reactions of the C-centered amino acid radicals with O2 and of the resultant AAOO with Hasc  allowed simultaneous observation of the course of both reactions at 360 nm (Fig. 5A). The rate constants of the oxidation of the Hasc  by the three AAOO were close to 107 M  1 s  1 (Table 2). This is not surprising, given the similar chemical structures of the parent molecules, which suggests that the value of 107 M  1 s  1 is likely to be generally representative of other free and protein-bound AAOO. Earlier measurements of the reactions of simple peroxyl radicals with Hasc  gave rate constants about 10  lower, except where the radicals were activated by the presence of chlorine [24,37,55]. Reaction rate constants of the peroxyl radicals of Gly, Ala, and Pro derivatives with TMPD were 5–10 times higher than the corresponding constants for the Hasc  reactions (Table 2), in agreement with previous studies with alkyl peroxyl radicals [55]. The knowledge of these constants combined with the stability of the TMPD þ generated (Fig. 6B) will allow the derivation of rate constants of the reactions of these AAOO with many other compounds by competitive kinetics, without the need for access to pulse radiolysis. Biological implications The well-documented loss of Hasc  in living organisms, including humans, subjected to oxidative stress poses the question of the nature of the responsible processes. Such processes are likely to be dominated by reactions between the Hasc  and an oxidizing radical, because in the generally reducing cell environment radicals are likely to constitute the principal species with sufficiently high reduction potentials and fast reaction kinetics. The submillimolar concentrations of Hasc  in most human tissues [38], together with a large variety of alternative radical targets, mean that oxidation of vitamin C will involve secondary radicals, generated in more abundant cell constituents. The protein C-centered radicals are the most common secondary species generated, with S- and N-centered species produced in smaller amounts, all having the capacity to react with the Hasc  [39]. The discussion of the potential biological chemistry based on kinetics can at present be applied only to homogeneous systems. This imposes a limit on the estimates of the relative significance of reactions involving protein radicals, Hasc  , and O2 in vivo, allowing one to distinguish between likely and unlikely reactions only in the aqueous compartments of cells, not in membranes or structures containing little water.

* **

**

** *

Scheme 1. Formation and reactions of protein radicals in vivo. Boxes marked

n

show protein repair, and those marked

nn

show protein damage.

1572

A.S. Domazou et al. / Free Radical Biology and Medicine 53 (2012) 1565–1573

In aerobic organisms, the probability of formation of peroxyl radicals is high. It has long been recognized that organic peroxides and their precursor peroxyl radicals are likely to be important intermediates in the pathway linking oxidative stress and its outcomes. The results of our study show that the oxidation of Hasc  by the peroxyl radicals generated in amino acids is a relatively rapid process and suggest its high probability of occurrence in vivo. Although we were limited in the selection of the amino acids, it is possible to apply these results to the likely course of reactions, which could count for much of the loss of Hasc  in cells and tissues under oxidative stress. In this, we assume that in vivo proteins are the primary target of radicals generated during the stress and that the initial reactions result in the formation of C-centered radicals on the amino acid residues [40]. The principal fates of these radicals are shown in Scheme 1, which is based on the finding that many of the PROS generated in oxidative stress have the capacity to oxidize proteins [40,41]. In the case of the HO, about 70% of the randomly generated radicals will attack the proteins, forming Pr. The relative probabilities of the subsequent reactions can be estimated from the relevant rate constants and tissue concentrations of the principal reactants, O2, Hasc  , and the –SH groups. In human tissues, the concentration of O2 is 30 mM [42], [Hasc  ] has an average value of about 750 mM [38], and that of RSH is close to 3 mM [43]. The approximate bimolecular reaction rate constants (all in M  1 s  1) of the reactions of the amino acid and/or protein C-centered radicals with O2, Hasc  , and RSH are 2  109 (this study), 6  107 [18], and  105 [56], respectively. This k value for the Hasc  reaction with Trp and Tyr assumes that the reactions of other C-centered AA have similar values of k, giving the probability ratios for reactions of the Pr with O2, Hasc  , and RSH in most human tissues of 1:0.75:0.005. Thus, there is a roughly equal chance of ‘‘repair’’ of Pr by Hasc  and the formation of PrOO (Reaction (2)), which can be reduced to PrOOH in reactions analogous to Reactions (10) and (11). It is important to stress that the ratios of the reaction probabilities will vary with the local concentrations of the reactants. The RSH compounds are unlikely to be an important route for Pr reduction, with the competition between O2 and Hasc  determining whether PrH or PrOOH will form ultimately. The most important determinants are the concentrations of the Hasc  , which typically vary from the average of 2.6 mM in the pituitary to 40 mM in the plasma [38]. At these extremes, the ratio for the probabilities of reactions of Pr with O2 and Hasc  is 1:2.6 for the pituitary gland and 1:0.02 for plasma, with different tissue-dependent ratios in between. For these two examples, the stable product of the reactions in the pituitary will be mainly PrH, whereas in plasma the PrOOH will predominate. Regardless of whether the Pr will react with O2 or Hasc  , either process constitutes biological damage. First, the antioxidant potential of the organism is depleted by the loss of Hasc  in Reaction (13) or (14): Pr þHasc  -PrHþasc 

(13)

PrOO þHasc  -PrOOHþ asc 

(14)

Our results suggest that these reactions can make a major contribution to the well-documented lowering of the levels of the Hasc  in living organisms under oxidative stresses. Second, Reaction (13) is likely to restore, at best, only part of the function of the protein, because the stereochemistry of the intermediate amino acid(s) radical can change. Finally, formation of the PrOOH can constitute a starting point for damage propagation, as demonstrated in many studies.

Acknowledgments This work was supported by Macquarie University, Sydney, Australia, and Eidgenossische Technische Hochschule, Zurich, Switzerland. References [1] Halliwell, B.; Gutteridge, J. Free Radicals in Biology and Medicine. Oxford: Clarendon Press; 2007. [2] Witenberg, B.; Kletter, Y.; Kalir, H. H.; Raviv, Z.; Fenig, E.; Nagler, A.; Halperin, D.; Fabian, I. Ascorbic acid inhibits apoptosis induced by X irradiation in HL60 myeloid leukemia cells. Radiat. Res. 152:468–478; 1999. [3] Li, X.; Hill, K.; Burk, R.; May, J. Selenium spares ascorbate and alphatocopherol in cultured liver cell lines under oxidant stress. FEBS Lett 508:489–492; 2001. [4] Vatassery, G.; Lai, J.; DeMaster, E.; Smith, W.; Quach, H. Oxidation of vitamin E and vitamin C and inhibition of brain mitochondrial oxidative phosphorylation by peroxynitrite. J. Neurosci. Res. 75:845–853; 2004. [5] Kume-Kick, J.; Rice, M. Estrogen-dependent modulation of rat brain ascorbate levels and ischemia-induced ascorbate loss. Brain Res 803:105–113; 1998. [6] Moor, E.; Shohami, E.; Kanevski, E.; Grigoriadis, N.; Symeonidou, C.; Kohen, R. Impairment of the ability of the injured aged brain in elevating urate and ascorbate. Exp. Gerontol. 41:303–311; 2006. [7] Buffinton, G.; Doe, W. Altered ascorbic acid status in the mucosa from inflammatory bowel disease patients. Free Radic. Res 22:131–143; 1995. [8] Foy, C. J.; Passmore, A. P.; Vahidassr, M. D.; Young, I. S.; Lawson, J. T. Plasma chain-breaking antioxidants in Alzheimer’s disease, vascular dementia and Parkinson’s disease. Q. J. Med. 92:39–45; 1999. [9] Locatelli, F.; Canaud, B.; Eckardt, K. U.; Stenvinkel, P.; Wanner, C.; Zoccali, C. Oxidative stress in end-stage renal disease: an emerging threat to patient outcome. Nephrol. Dial. Transplant. 18:1272–1280; 2003. [10] Misso, N. L.; Brooks-Wildhaber, J.; Ray, S.; Vally, H.; Thompson, P. J. Plasma concentrations of dietary and nondietary antioxidants are low in severe asthma. Eur. Respir. J. 26:257–264; 2005. [11] Davies, M. J. The oxidative environment and protein damage. Biochim. Biophys. Acta 1703:93–109; 2005. [12] Gebicki, S; Gebicki, J. M. Formation of peroxides in amino acids and proteins exposed to oxygen free radicals. Biochem. J. 289:743–749; 1993. [13] Gieseg, S.; Duggan, S.; Gebicki, J. M. Peroxidation of proteins before lipids in U937 cells exposed to peroxyl radicals. Biochem. J. 350:215–218; 2000. [14] Du, J.; Gebicki, J. M. Proteins are major initial cell targets of hydroxyl free radicals. Int. J. Biochem. Cell Biol. 36:2334–2343; 2004. [15] Gebicki, J. M. The role of proteins in biological damage induced by oxidative stress. In: Pietzsch, J., editor. Protein Oxidation and Disease. Trivandrum: Research Signpost; 2006. p. 7–38. [16] Hoey, B. M.; Butler, J. The repair of oxidized amino acids by antioxidants. Biochim. Biophys. Acta 791:212–218; 1984. [17] 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. 33:383–391; 2000. [18] Domazou, A. S.; Koppenol, W. H.; Gebicki, J. M. Efficient repair of protein radicals by ascorbate. Free Radic. Biol. Med. 46:1049–1057; 2009. [19] Cudina, I.; Josimovic, L. The effect of oxygen on the radiolysis of tyrosine in aqueous solutions. Radiat. Res 109:206–215; 1987. [20] 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. 67:229–237; 1997. [21] Willson, R. Organic peroxy free radicals as ultimate agents in oxygen toxicity. In: Sies, H., editor. Oxidative Stress. London: Academic Press; 1985. p. 41–72. [22] Fujita, S.; Steenken, S. Pattern of OH radical addition to uracil and methyland carboxyl-substituted uracils: electron transfer of OH addicts with N,N,N0 ,N0 -tetramethyl phenylene diamine and tetranitromethane. J. Am. Chem. Soc 103:2540–2545; 1981. [23] Buxton, G.; Stuart, C. Re-evaluation of the thiocyanate dosimeter for pulse radiolysis. J. Chem. Soc. Trans. Faraday Soc 91:279–281; 1995. [24] von Sonntag, C. The Chemical Basis of Radiation Biology. London: Taylor & Francis; 1987. [25] Buxton, G.; Greenstock, C.; Helman, W.; Ross, A. 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 17:513–886; 1988. [26] Xu, G.; Takamoto, K.; Chance, M. R. Radiolytic modification of basic amino acid residues in peptides: probes for examining protein–protein interactions. Anal. Chem. 75:6995–7007; 2003. [27] Nauser, T.; Casi, G.; Koppenol, W. H.; Schoneich, C. Reversible intramolecular hydrogen transfer between cysteine thiyl radicals and glycine and alanine in model peptides: absolute rate constants derived from pulse radiolysis and laser flash photolysis. J. Phys. Chem. B 112:15034–15044; 2008. [28] Bielski, B. H. J.; Allen, A. O. Absorption spectrum and decay rate of the ascorbic acid radical. J. Am. Chem. Soc. 92:3793–3794; 1970. [29] Wolfenden, B.; Willson, R. Radical-cations as reference chromogens in kinetic studies of one-electron transfer reactions: pulse radiolysis studies of

A.S. Domazou et al. / Free Radical Biology and Medicine 53 (2012) 1565–1573

[30]

[31]

[32] [33] [34] [35]

[36]

[37] [38]

[39] [40] [41]

2,20 -azinobis-(3-ethylbenzthiazoline-6-sulphonate). J. Chem. Soc. Perkin Trans 2:805–812; 1982. Hawkins, C. L.; Davies, M. J. EPR studies on the selectivity of hydroxyl radical attack on amino acids and peptides. J. Chem. Soc. Perkin Trans 2:2617–2622; 1998. Rustgi, S.; Joshi, A.; Moss, H.; Riesz, P. E.s.r. of spin-trapped radicals in aqueous solutions of amino acids: reactions of the hydroxyl radical. Int. J. Radiat. Biol. 31:415–440; 1977. Draganic, I. G.; Draganic, Z. D. The Radiation Chemistry of Water. New York: Academic Press; 1971. 92. Samuni, A. Precursors of the metal-complexed hydroperoxyl radical. J. Phys. Chem. 76:2207–2213; 1972. Dorfman, L. -M.; Adams, G. E. National Standard Reference Data System, 46. Washington, DC: Natl. Bur. of Standards; 1973. Burgess, V. A.; Easton, C. J.; Hay, M. P. Selective reaction of glycine residues in hydrogen atom transfer from amino acid residues. J. Am. Chem. Soc. 111:1047–1052; 1989. Shono, T.; Matsumura, Y.; Tsubata, K.; Sugihara, Y.; Yamane, S.; Kanasawa, T.; Aoki, T. Electroorganic chemistry. 60. Electroorganic synthesis of enamides and enecarbamates and their utilization in organic synthesis. J. Am. Chem. Soc. 104:6697–6703; 1982. Neta, P.; Simic, M; Hayon, E. Pulse radiolysis of aliphatic acids in aqueous solutions. J. Phys. Chem. 73:4207–4213; 1969. Hayon, E.; Ibata, T.; Lichtin, N. N.; Simic, M. Sites of attack of hydroxyl radicals on amides in aqueous solution. II. Effects of branching alpha to carbonyl and to nitrogen. J. Am. Chem. Soc. 93:5388–5394; 1971. Hayon, E.; Ibata, T.; Lichtin, N. N.; Simic, M. M. Sites of attack of hydroxyl radials on amides in aqueous solution. J. Am. Chem. Soc. 92:3898–3903; 1970. Rao, P. S.; Hayon, E. Reaction of hydroxyl radials with oligopeptides in aqueous solutions. J. Phys. Chem. 79:109–115; 1975.  and OH Simic, M.; Neta, P.; Hayon, E. Selectivity in the reactions of eaq radicals with simple peptides in aqueous solution: optical absorption spectra of intermediates. J. Am. Chem. Soc 92:4763–4768; 1970.

1573

[42] Easton, C. J. Free radical reactions in the synthesis of alpha-amino acids and their derivatives. Chem. Rev 97:53–82; 1997. [43] Garrison, W. Radiation-induced reactions of amino acids and peptides. Radiat. Res. Rev 3:305–326; 1972. [44] Trelstad, R. L.; Lawley, K. R.; Holmes, L. B. Nonenzymatic hydroxylations of proline and lysine by reduced oxygen derivatives. Nature 289:310–312; 1981. [45] Floyd, R. A.; Nagy, I. Formation of long-lived hydroxyl free radical adducts of proline and hydroxyproline in a Fenton reaction. Biochim. Biophys. Acta 790:94–97; 1984. [46] Uchida, K.; Kato, Y.; Kawakishi, S. A novel mechanism for oxidative cleavage of prolyl peptides induced by the hydroxyl radical. Biochem. Biophys. Res. Commun. 169:265–271; 1990. [47] Stadtman, E. R.; Levine, R. L. Free radical-mediated oxidation of free amino acids and amino acid residues in proteins. Amino Acids 25:207–218; 2003. [48] Packer, J. E.; Willson, R. L.; Bahnemann, D.; Asmus, K. -D. Electron transfer reactions of halogenated aliphatic peroxyl radicals. J. Chem. Soc. Perkin Trans. 2:296–299; 1980. [49] Neta, P.; Huie, R. E.; Mosseri, S.; Shastri, L. V.; Mittal, J. P.; Maruthamuthu, P.; Steenken, S. J. Rate constants for reduction of substituted methylperoxyl radicals by ascorbate and N,N,N0 ,N0 -tetramethyl-p-phenylenediamine. J. Phys. Chem. 93:4099–4104; 1989. [50] Hornig, D. Distribution of ascorbic acid, metabolites and analogues in man and animals. Ann. N. Y. Acad. Sci. 258:103–118; 1975. [51] von Sonntag, C. Free-Radical-Induced DNA Damage and Its Repair. Berlin: Springer-Verlag; 2006. [52] Davies, M. J.; Dean, R. T. Radical-Mediated Protein Oxidation. Oxford: Oxford Univ. Press; 1997. [53] Gebicki, J. M. Protein hydroperoxides as new reactive oxygen species. Redox Rep. 3:99–110; 1997. [54] Boyle 3rd J. Microcirculatory hematocrit and blood flow. J. Theor. Biol. 131:223–229; 1988. [55] Jocelyn, P. C Biochemistry of the SH Group. New York: Academic Press; 1972. [56] Nauser, T.; Koppenol, W. H.; Gebicki, J. M. The kinetics of oxidation of GSH by protein radicals. Biochem. J. 392:693–701; 2005.