Chem. Res. Toxicol. 2001, 14, 1453-1464
1453
Absolute Rate Constants for the Reaction of Hypochlorous Acid with Protein Side Chains and Peptide Bonds David I. Pattison and Michael J. Davies* Contribution from The Heart Research Institute, 145 Missenden Road, Camperdown, Sydney, NSW 2050, Australia Received July 30, 2001
Hypochlorous acid (HOCl) is a potent oxidant, which is produced in vivo by activated phagocytes. This compound is an important antibacterial agent, but excessive or misplaced production has been implicated in a number of human diseases, including atherosclerosis, arthritis, and some cancers. Proteins are major targets for this oxidant, and such reaction results in side-chain modification, backbone fragmentation, and cross-linking. Despite a wealth of qualitative data for such reactions, little absolute kinetic data is available to rationalize the in vitro and in vivo data. In this study, absolute second-order rate constants for the reactions of HOCl with protein side chains, model compounds, and backbone amide (peptide) bonds have been determined at physiological pH values. The reactivity of HOCl with potential reactive sites in proteins is summarized by the series: Met (3.8 × 107 M-1 s-1) > Cys (3.0 × 107 M-1 s-1) . cystine (1.6 × 105 M-1 s-1) ≈ His (1.0 × 105 M-1 s-1) ≈ R-amino (1.0 × 105 M-1 s-1) > Trp (1.1 × 104 M-1 s-1) > Lys (5.0 × 103 M-1 s-1) . Tyr (44 M-1 s-1) ≈ Arg (26 M-1 s-1) > backbone amides (10-10-3 M-1 s-1) > Gln(0.03 M-1 s-1) ≈ Asn (0.03 M-1 s-1). The rate constants for reaction of HOCl with backbone amides (peptide bonds) vary by 4 orders of magnitude with uncharged peptide bonds reacting more readily with HOCl than those in a charged environment. These kinetic parameters have been used in computer modeling of the reactions of HOCl with human serum albumin, apolipoprotein-A1 and free amino acids in plasma at different molar excesses. These models are useful tools for predicting, and reconciling, experimental data obtained in HOCl-induced oxidations and allow estimations to be made as to the flux of HOCl to which proteins are exposed in vivo.
Introduction Hypochlorous acid is a potent oxidant, which is produced in vivo by activated phagocytes. It is generated by release of the heme enzyme myeloperoxidase, which catalyzes the reaction of H2O2 with Cl- ions (1-3). The pKa of hypochlorous acid is 7.59 (4), thus at physiological pH values, it exists in equilibrium with the anion -OCl. This mixture will be referred to as HOCl1 throughout this paper. HOCl has important antibacterial properties, but excessive or misplaced production of HOCl has been implicated in several diseases, including atherosclerosis, inflammatory diseases (e.g., arthritis), and some cancers associated with inflammation (1-3). Studies on protein oxidation by HOCl have shown that side-chain modification, protein fragmentation, and crosslinking/aggregation can occur in vitro (5-12). It is generally accepted that the sulfur-containing residues, Cys and Met, react most rapidly with HOCl, but that the side chains of Lys, His, Trp, and Tyr, and the R-amino * To whom correspondence should be addressed: Phone: 61-2-95503560. Fax: 61-2-9550-3302. Email:
[email protected]. 1 Abbreviations: apo-A1, human apolipoprotein-A1; BSA, bovine serum albumin; 3-Cl-HPPA, 3-(3-chloro-4-hydroxyphenyl)propionic acid; 3-Cl-Tyr, 3-chlorotyrosine; 3,5-Cl2-Tyr, 3,5-dichlorotyrosine; DTDPA, 3,3′-dithiodipropionic acid; HOCl, the physiological mixture of hypochlorous acid and its anion; HPAA, 4-hydroxyphenylacetaldehyde; HPPA, 3-(4-hydroxyphenyl)propionic acid; HSA, human serum albumin; IPA, 3-indolepropionic acid; MCD, monochlorodimedon.
groups of amino acids and peptides, are also modified (5, 7-11, 13-18). The reaction of HOCl with Cys produces disulfides and oxy acids (5, 19-22), with sulfenyl chlorides as likely intermediates (23, 24). Reaction with Met produces the sulfoxide (5, 22, 25-27). Unstable chloramines (RNHCl and RR′NCl) are generated on reaction of HOCl with Lys side chains (28, 29). These species can decompose to generate aldehydes and ketones (29, 30) or nitrogen-centered radicals; the latter can undergo further reactions (10-12). Further reaction of monochloramines with excess HOCl can lead to the formation of dichloramines (29), though these species are unlikely to play a role in protein damage in most biological situations. The reaction of HOCl with Tyr residues can result in the formation of 3-chlorotyrosine (3-Cl-Tyr) and 3,5dichlorotyrosine (3,5-Cl2-Tyr). The formation of these materials has been employed as a marker for HOCl-, and myeloperoxidase-, catalyzed reactions in vitro and in vivo (14, 15, 18, 31, 32). The products formed on reaction of HOCl with His have not been determined in any detail. With Trp, 2-oxoindole or 2-oxoindolone derivatives are generated (25, 33), though the mechanisms of formation of these materials are unclear. It has recently been reported that HOCl oxidation of apolipoprotein-A1 (apoA1) results in the chlorination of the Phe (27), though this residue is not generally recognized as a major target for HOCl oxidation.
10.1021/tx0155451 CCC: $20.00 © 2001 American Chemical Society Published on Web 09/11/2001
1454
Chem. Res. Toxicol., Vol. 14, No. 10, 2001
Pattison and Davies
Despite the wealth of qualitative data available on the effects of HOCl on proteins, little absolute kinetic data is available. Comparative rates of reaction of HOCl with selected free amino acids have been determined by use of monochlorodimedon (MCD), and this yielded a reactivity scale of Cys > Met > Cystine > His > Ser > Leu (5). These data indicate that Cys and Met are ca. 100 times more reactive than the other residues tested, but in most cases, these amino acids contained multiple potential sites of reaction (side-chain and R-amino group), and hence these comparative rates cannot be assigned to particular processes. Absolute rate constants have been determined for the reaction of HOCl with a few free amino acids (17, 34). The second-order rate constants determined, at physiological pH values, are in the region of 104-105 M-1 s-1 and are for reaction of HOCl with the R-amino position of amino acids (Gly, Ala, Val) with no reactive side chains (34). Similar studies were carried out with di- and tri-peptides, but the R-amino groups were not protected, thus the majority of the reactivity is likely to be at these sites, with k2 ≈ 105 M-1 s-1 (35). Recently, absolute second-order rate constants have been reported for reaction with Cys and Met at high pH (10-14), with k2 ≈ 105-106 M-1 s-1 for Cys, and k2 ≈ 104-105 M-1 s-1 for Met (17). These values are highly pH dependent, due to the effect of the pKas of HOCl and the substrate (reactions 1 and 2) (17).
RS- + HOCl f RSCl + OHH2O
RS- + OCl- 98 RSCl + 2OH-
(1) (2)
The purpose of the work described here was to determine absolute rate constants at physiological pH (7.4) for the reaction of HOCl with the various side chains of proteins, and backbone amide groups. The resulting kinetic data have been used to predict the reactivity of HOCl with the plasma proteins, human serum albumin (HSA), and apo-A1, and free amino acids. The predictions from these models have been compared with experimental data.
Experimental Procedures Materials. Amino acids, derivatives and model compounds were from Sigma/Aldrich and used without further purification. The peptides, N-acetyl-Ala, N-acetyl-Leu-OMe, N-acetyl-(Ala)2, N-acetyl-(Ala)2-OMe, N-acetyl-(Ala)3, N-acetyl-(Ala)3-OMe, and cyclo(Asp)2 were from Bachem and used as received. All reactions were performed in 0.1 M Chelex treated phosphate buffer (pH 7.4) solutions, prepared using Milli Q treated water. The pH values of solutions were adjusted, where necessary, to pH 7.4 using 1.0 M HCl, or 1.0 M NaOH. For the majority of the kinetic experiments, HOCl was kept as the limiting reagent, with a greater than 2-fold excess of substrate (typically 0.51.0 mM HOCl, 5.0-20.0 mM substrate). With highly absorbing substrates [e.g., 3-indolepropionic acid (IPA)] the concentrations were reduced up to 200-fold (e.g., 25 µM HOCl, 50-150 µM substrate). The excess of reagent was employed to reduce the kinetics to pseudo-first-order behavior for ease of analysis and to reduce the likelihood of secondary reactions including dichloramine formation. Stopped-Flow Studies. Two stopped-flow systems were used depending on the time-scale being studied. For rapid kinetics (10 ms to 10 s time scale), an Applied Photophysics SX.18MV system controlled by an Acorn 5000 computer was used. Kinetic data were acquired with a monochromator and photomultiplier detection, and a linear timebase (deadtime, ∼2.5 ms). For slower reactions (over the 2-300 s time scale), a Hi-
Figure 1. Decay of MCD absorbance when HOCl (20 µM) was reacted with MCD (20 µM) at pH 7.4 and 22 °C. The data represents the decay over 100 ms, with a 5 ms interval between spectra. Tech SFA 20 attachment (deadtime, ∼0.25 s) was used in conjunction with a PC-controlled Perkin-Elmer Lambda 40 UVvis spectrometer (slit width, 4.0 mm; response, 0-0.2 s; data interval, 0.05-0.5 s). For both systems, the temperature was controlled (at either 10 or 22 °C) by circulating water from a thermostated water bath. In the case of the SFA 20 system, the cuvette holder was additionally thermostated by means of a Peltier block. Kinetic traces were measured at 10 nm intervals between 220 and 350 nm (0.1 M phosphate buffer baseline) and combined to give time-dependent spectral data. The data were processed by global analysis methods, either using Pro-Kineticist (Applied Photophysics) or Specfit (Version 3.0.15, Spectrum Software Associates) software. UV-Vis Spectroscopy. UV-vis spectroscopy was undertaken on a Perkin-Elmer Lambda 40 spectrometer. Spectra were typically acquired (relative to a 0.1 M phosphate buffer baseline) between 220 and 320 nm (1-2 nm intervals), with a time interval of 1-30 min depending on the reaction time scale. The cuvettes were thermostated to 22 °C using a Peltier block. Data were imported into Specfit software, and the kinetics processed by global analysis techniques. Statistics. All the second-order rate constants reported are averages of at least four separate determinations, and errors are specified as 95% confidence limits. Kinetic Modeling of HOCl Reactivity with Proteins and Plasma Amino Acids. The modeling of HSA, apo-A1, and plasma amino acid reactivity with HOCl was performed with Specfit software. A series of individual reactions describing the full reactivity of the proteins was used as the model, and the software underwent 100 iterations to yield predicted reactant and product concentrations after specified time intervals.
Results Kinetics of HOCl Reaction with Monochlorodimedon (MCD). Absolute rate constants for reaction of HOCl with MCD were determined to allow comparison of the absolute rate constants determined here with the relative rates determined previously (5). MCD exhibits a strong absorption at 290 nm, thus kinetics were determined by monitoring the loss of MCD absorbance (Figure 1) at low reactant concentrations (10 µM HOCl, 10-50 µM MCD). The rates were determined at both 10 and 22 °C to provide a means of predicting rates at 22 °C from data measured at 10 °C. At both temperatures, satisfactory second-order fits to the data were obtained at a variety of MCD concentrations, yielding k2 (2.2 ( 0.9) × 106 M-1 s-1 at 10 °C, and k2 (3.6 ( 0.7) × 106 M-1 s-1 at 22 °C, with a ratio of k2(22 °C)/k2(10 °C) of 1.64. Attempts to determine the second-order rate of reaction between HOCl and MCD at 37 °C were unsuccessful, as the observed reaction rates were too fast to obtain reproducible data.
Rate Constants for HOCl with Protein Side Chains
Chem. Res. Toxicol., Vol. 14, No. 10, 2001 1455
Table 1. Second Order Rate Constants (with 95% confidence limits) Determined (at pH 7.2-7.4) for the Reactions of HOCl with r-Amino Groups and Side-Chain Model Compounds
1456
Chem. Res. Toxicol., Vol. 14, No. 10, 2001
Pattison and Davies
Table 1 (Continued)
a Rate calculated for 22 °C using data obtained at 10 °C and the ratio of the second-order rate constants for monochlorodimedon as a reference. ND, rate not determined at 10 °C.
Kinetics of HOCl Reaction with Amino Acid Side Chains. Initial studies employed amino acids with aliphatic side chains (Gly, Ala, Val, Ser), thereby limiting reaction to the R-amino group. Due to the rapid kinetics observed, these studies were carried out at 10 °C to slow the reactions, allowing more accurate kinetic data to be determined. Over a period of 100 ms with 0.5 mM HOCl, and 2.5-10 mM amino acid, the absorbance due to -OCl at 290 nm decayed completely, and a concomitant rise in absorbance was observed at 250 nm. The new peak has been assigned to the monochloramine formed at the R-amino position (29). The second-order rates determined for each amino acid at 10 °C are listed in Table 1, together with the predicted rates at 22 °C, calculated using the ratio of the rates determined for MCD with HOCl as a conversion factor. The reactivity of HOCl with the Lys side chain was investigated with three model compounds; -amino-ncaproic acid, N-acetyl-Lys, and tert-butylamine (see Table
1 for structures). The decay of the -OCl peak was monitored at 290 nm, with formation of the chloramine peak at 250 nm. Rate constants were determined at 10 °C (Table 1), as the rates were too fast at 22 °C; rate constants for the latter temperature were calculated as described above. Reaction with the imidazole ring of the His side chain was determined using 4-imidazoleacetic acid, and Nacetyl-His where the R-amino group is blocked. For both compounds, the -OCl absorbance disappeared rapidly, even at 10 °C, and a new absorbance band appeared with rising intensity in the UV region of the spectrum. The time-dependent spectra were fitted satisfactorily by global analysis, and yielded the rates shown in Table 1. The effect of pH on the reactivity of the His side-chain was investigated as its pKa is known to vary in proteins and is close to physiological pH (pKa ca. 6.0). The pH was varied in the range from 5.9 to 8.8, and the second-order rate constants determined at 10 °C (Figure 2).
Rate Constants for HOCl with Protein Side Chains
Chem. Res. Toxicol., Vol. 14, No. 10, 2001 1457
reaction with the phenolic ring (Table 1). N-Acetyl-Tyr yielded similar data (Table 1). k1
k2
HOCl + HPPA 98 Intermediate 98 Cl-HPPA (4)
Figure 2. pH dependence of the second-order rate constants for the reaction of HOCl with the His side-chain model, 4-imidazole acetic acid.
It has recently been reported that the reaction of HOCl with Phe-containing peptides yields chlorinated Phe residues (27). Reaction of HOCl with 3-phenylpropionic acid, a model for the Phe side chain, was investigated (1.0 mM HOCl, 2.0 mM 3-phenylpropionic acid, 22 °C), but no decay of the -OCl absorbance (at 290 nm) was seen over 15 h. Direct reaction of HOCl with the Phe side chain is therefore slow. The reactivities of other residues with nitrogen containing side chains were also studied. The reactivity of the Arg side chain was investigated using ethyl guanidine and N-acetyl-Arg-OMe. The reaction kinetics were monitored directly at 22 °C, with substrate concentrations between 2.5 and 10 mM. In both cases, the decay of -OCl absorbance at 290 nm was accompanied by an increase in absorbance at shorter wavelengths, but over time this new absorbance also decayed. The kinetic profiles obtained were most accurately fitted by the mechanism shown in reaction 5; the nature of the intermediate and product(s) are the subject of current studies; the resulting second-order rate constants are given in Table 1. Studies with acetamidine gave similar behavior and kinetics, though the rate constant was ca. 5-fold larger (Table 1). k1
Figure 3. Kinetics observed (at 280 nm) for the reaction of HOCl (25 µM) with the Trp side-chain model, IPA (75 µM), showing the initial consumption of HOCl, and subsequent reactions.
HOCl + Arg side chain 98
Reaction of HOCl with the Trp side chain was studied using IPA and N-acetyl-Trp. Due to the strong absorbance of the indole ring at short wavelengths, these experiments were performed with low reactant concentrations (25 µM HOCl, 50-150 µM substrate), with the loss in absorbance monitored at 22 °C. The resulting kinetics were complex (Figure 3). A rapid initial decay was observed at 280 nm, which varied linearly with the concentration of the model compound present, confirming direct reaction of HOCl with the substrate; the secondorder rate constants obtained are given in Table 1. The subsequent absorbance changes were fitted with the mechanism in reaction 3 to give first-order rate constants of ca. 0.05 and 0.005 s-1, respectively. The identities of the species responsible for these subsequent complex kinetics are the subject of a further publication; products A or B may be the previously detected 2-oxoindole or 2-oxoindolone derivatives (25, 33).
Reaction with amide side chains (e.g., Gln and Asn) was modeled using the aliphatic amides, propionamide, 3-methylpropionamide (isobutyramide) and 3,3-dimethylpropionamide (trimethylacetamide). Reaction of these compounds (8.0-25 mM) with HOCl (1.0 mM) was very slow. The -OCl absorbance (270-320 nm) decayed with pseudo-first-order kinetics and yielded the second-order rate constants in Table 1. At shorter wavelengths (220260 nm) more complex kinetics were observed, with a rise and subsequent decay in absorbance; these changes are assigned to the formation and decay of chloramide [RC(O)NHCl] species. 3,3′-Dithiodipropionic acid (DTDPA) was studied as a model for cystine. The kinetics were investigated with 1.0-2.5 mM DTDPA and 0.5 mM HOCl, due to the strong absorbance of DTDPA at 255 nm. On the millisecond time scale, changes in absorbance were observed that were consistent with consumption of HOCl and formation of a new species. The rate of HOCl decay varied with the concentration of DTDPA and yielded the second-order rate constant (Table 1). This reaction was also investigated over longer time scales (up to 60 s), whereupon a relatively slow growth was observed at 235 nm. The rate of this subsequent growth (19 ( 1 s-1) was invariant at different DTDPA concentrations indicating that this growth is due to a further reaction of an uncharacterized initial product. Kinetics of HOCl Reaction with Backbone Amides. As the concentration of backbone amide groups in a protein is higher than that of any individual side chain, direct reaction of HOCl with backbone amide groups may be significant under some circumstances. Reaction with model peptides was therefore studied by monitoring the
k1
k2
HOCl + IPA 98 Chloramine 98 k3
Product A 98 Product B (3) The Tyr side-chain model compound, 3-(4-hydroxyphenyl)propionic acid (HPPA), was reacted (at 22 °C) with HOCl at low concentrations (200 µM HOCl, 0.35-1.0 mM HPPA), due to the strong absorbance of the phenolic ring. Over a time period of 4 min, growths in absorbance were detected at wavelengths (240 and 300 nm) on either side of the HPPA peak. Comparison with the spectra for authentic 3-Cl-Tyr suggested that the ultimate product is 3-Cl-HPPA. The kinetics were fitted with the mechanism in reaction 4, and yielded a low rate constant for
k2
Intermediate 98 Products (5)
1458
Chem. Res. Toxicol., Vol. 14, No. 10, 2001
Pattison and Davies
Table 2. Second Order Rate Constants (with 95% confidence limits) for the Reaction of HOCl with Models for Backbone Amide Groups at 22 °C and pH 7.2-7.4; All Rates Are Expressed Per Amide Group in the Substrate
a
substrate
k2 (22 °C) (M-1 s-1)
cyclo (Gly)2 cyclo (Ala)2 cyclo (Ser)2 cyclo (Asp)2 N-acetyl-Ala N-acetyl-Leu-OMe N-acetyl-(Ala)2 N-acetyl-(Ala)2-OMe N-acetyl-(Ala)3 N-acetyl-(Ala)3-OMe
25 ( 5a 8.2 ( 2.1a 16 ( 3a 1.9 ( 0.9 0.0012 ( 0.001 0.015 ( 0.007 0.008 ( 0.004 0.06 ( 0.015 0.023 ( 0.018 0.11 ( 0.02
Rates calculated from data acquired at 10 °C (see text).
absorbance due to -OCl at 290 nm. A new absorbance with rising intensity in the UV region was observed; this absorbance band has been assigned to chloramide [RC(O)NClR′] formation. The second-order rate constants for these reactions varied by 4 orders of magnitude (from 10-3 to 10 M-1 s-1 per amide bond), depending on the environment (Table 2). In general, reaction was faster with uncharged substrates [e.g., cyclo(Gly)2, N-acetylLeu-OMe], than charged analogues [e.g., cyclo(Asp)2, N-acetyl-Ala]. Complex kinetics were observed at shorter wavelengths, due to the subsequent decay of the initially formed chloramides. Reaction of HOCl with Sulfur-Containing Residues. It was not possible to determine absolute rate constants for reaction of Cys and Met with HOCl, at pH 7.4, as these reactions were too fast for accurate measurement. Previous competition kinetic data obtained using MCD (5) have been used in conjunction with the absolute rate constant determined here for reaction of HOCl with MCD, to predict absolute second-order rate constants for these reactions (Table 3). Armesto et al. have previously investigated the reaction of HOCl with Cys and Met at high pH values (11-14) (17). This group analyzed the data obtained using an expression (eq 6; KHOCl, KSH, ionization constants for HOCl and the Cys side chain respectively; k(RS-+HOCl), k(RS-+-OCl), rate constants for processes in reactions 1 and 2, respectively) incorporating terms for the ionization constant of each ionizable group present, and the H+ concentration (17). However, the pKa values used by this group are inconsistent with some literature values (36). Reanalysis of the data with alternative pKa values [HOCl pKa 7.59; Cys(-SH) pKa 8.37] yielded revised rates for reactions 1 and 2 (Table 3). These results allow a rate constant to be estimated for reaction at pH 7.4 by extrapolation of the data using the fitting equation (eq 6, Table 3). A
Figure 4. pH dependence of the second-order rate constants for reaction of HOCl with Cys. The solid line represents the best fit to the data using the equation reported in (17) with pKa(SH) ) 8.37 and pKa(HOCl) ) 7.59.
similar approach has been employed to estimate a rate constant for reaction with Met at pH 7.4.
k2 )
[
k(RS-+HOCl)[H+] +
KHOCl + [H ]
+
]
k(RS-+-OCl)[H+] +
KSH
KHOCl + [H ] KSH + [H+]
(6)
The studies of Armesto et al on the reaction of Cys with HOCl (17), have been replicated and extended to provide further data in the low pH region. The observed secondorder rates of reaction of HOCl with Cys were determined at varying pH values (Figure 4) using the equation derived by Armesto et al. (17), with the revised pKa values given above. The resulting rate constants are given in Table 3. Kinetic Modeling of Protein Reactivity with HOCl. The absolute rate constants determined above have been used to generate a computational model (Table 4) for the reactivity of HOCl with proteins, and this has been applied to two human plasma proteins: HSA and apoA1. The amino acid compositions of these proteins have been taken to determine the relative abundance of the reactive sites in these proteins (Table 4). The reactivity of the various protein side chains and backbone sites with HOCl have been calculated with different molar excesses of HOCl, ranging from 0.16- to 170-fold. The rate constant for modeling the backbone amide reactivity was taken at the upper limit of those observed experimentally (k 10 M-1 s-1). This allows a maximal estimate of protein backbone attack, and represents a worse case scenario for direct backbone cleavage by HOCl in proteins. The results of these modeling studies have been expressed as the proportion of HOCl reacting with each residue (Figures 5a and 6a) to provide information as which sites the majority of HOCl reacts with at different ratios. This
Table 3. Predicted Second Order Rate Constants at pH 7.4 and 22 ˚C for the Reaction of HOCl with Cys and Met Residues by Different Methods substrate Cysa Cysb Cysc Cysd Meta Metb Metc
pKa(HOCl)
pKa(-SH)
k (RS- + HOCl) (M-1 s-1)
k(RS- + OCl-) (M-1 s-1)
7.26 7.59 7.59
8.15 8.37 8.37
1.2 × 109 4.0 × 108 3.3 × 108
1.9 × 105 2.2 × 105 5.5 × 105
7.26 7.59
k(Met + HOCl) (M-1 s-1)
k2 (pH 7.4) (M-1 s-1)
8.7 × 108 6.3 × 107
1.8 × 108 7.6 × 107 2.4 × 107 3.2 × 107 1.2 × 108 3.7 × 108 3.8 × 107
a Using relative rates to MCD from ref 5. b Using absolute kinetic data measured at high pH and analysis from ref 17. c Using absolute kinetic data measured at high pH from ref 17 and reanalysis with different pKa values. d Using absolute kinetic data measured and analyzed at high pH in these studies.
Rate Constants for HOCl with Protein Side Chains
Chem. Res. Toxicol., Vol. 14, No. 10, 2001 1459
Table 4. Parameters Used for Modeling the Reactivity of HOCl with HSA, Apo-A1, and Free Amino Acids in Plasmaa residue
HSA
apo-A1
His Lys Arg Asn Gln Cys Met cystine Tyr Trp terminal amine backbone
16 59 24 17 20 1 6 17 18 1 1 583
5 21 16 5 16 0 3 0 7 4 1 242
plasma amino acids (µM)
k2 (M-1 s-1)
89 167
1.0 × 105 5.0 × 103 26 0.03 0.03 3.0 × 107 3.8 × 107 1.6 × 105 44 1.1 × 104 1.0 × 105 10
33 24 42 54b 45 2584
a The numbers of reactive side-chain residues and backbone amides present in each protein are given, together with the second order rate constants used for the modeling, and the plasma concentrations of the individual reactive groups from free amino acids. b The rate constant used for Tyr in the plasma amino acid modeling was that for reaction at the R-amino group (i.e., 1.0 × 105 M-1 s-1), not direct ring chlorination.
Figure 6. Predicted reactivity of various molar excesses of HOCl with apo-A1 using the model described in Table 4, showing (a) the proportion of HOCl reactivity at each residue and (b) the percentage of each residue remaining. Abbreviations: Term, terminal amino group; Back, backbone amides.
Figure 5. Predicted reactivity of various molar excesses of HOCl with HSA using the model described in Table 4, showing (a) the proportion of HOCl reactivity at each residue and (b) the percentage of each residue remaining. Abbreviations: -SS-, cystine; Term, terminal amino group; Back, backbone amides.
varies with the HOCl:protein ratio, as reaction of HOCl with the less reactive residues becomes more prevalent as the more reactive residues are consumed. These results are also expressed as the percentage of the initial residue concentration remaining (Figures 5b and 6b), thereby providing information on the relative rates of depletion of each type of residue. The results obtained with HSA (Figure 5a) showed that at low molar excesses of HOCl (98%) is consumed by Cys and Met residues. With larger excesses of HOCl (8:1 and 17: 1) the proportion of HOCl consumed by the Met and Cys
residues reduces, with the remainder of the HOCl increasingly consumed by cystine, and the His and Lys residues. At HOCl excesses of 67:1 and 170:1, a major proportion of the HOCl reacts with Lys (up to 37%), with lesser amounts reacting at cystine (25%), His (25%), and Met (9%) residues, and low amounts reacting at Trp and the terminal amine group. Reaction with backbone amides (peptide bonds) becomes significant at these excesses, with 0.9% of the HOCl reacting at the backbone at a ratio of 67:1; this figure increases at higher ratios. The data on the proportion of each residue remaining at various HOCl:HSA ratios (Figure 5b) predicts that with an 8-fold excess of HOCl, the Met and Cys residues are fully depleted, but reaction at the other side chains is minimal. At higher ratios, the disulfide bonds, His, and terminal amine groups are oxidized, followed by Trp and Lys. Oxidation of Tyr and Arg residues, and backbone amides, become important only at high HOCl:proteins ratios, once reactive side-chain residues have been consumed. Apo-A1 (Figure 6a) shows different behavior to HSA at low HOCl excesses as this protein does not contain Cys residues or disulfide bonds. At low HOCl concentrations ( 99% of the HOCl is predicted to react with the Met residues. At increasing molar excess of HOCl (up to 17:1), the majority (>80%) of the HOCl reacts with the Met, His, and Lys residues, with the extent of reaction at His and Lys increasing at greater HOCl excesses. At a ratio of 17:1, a significant proportion of the HOCl is consumed by the terminal amine group ( Trp > Lys > Tyr ≈ Arg > backbone amides > Asn ≈ Gln. This order of reactivity is consistent with the limited, indirect, data reported previously (5) for free amino acids, where cystine > His > Ser > Leu. Peskin et al. have also recently reported that reaction with thiols is fast (k > 107 M-1 s-1) and only modestly affected by substrate structure, although these data were obtained by competitive kinetics rather than direct measurement (38). The data obtained previously by Winterbourn (5) did not allow information to be obtained as to the site of reaction with cystine and His (i.e., R-amino versus disulfide bond or imidazole ring, respectively). The present study shows that both the side chain and the R-amino sites contribute to the observed reactivity, with the free amino acids, with a slight preference for the disulfide bond in the case of cystine. The reactivity (k ca. 105 M-1 s-1; see also ref 41) of the disulfide bond, which has not been previously considered as a major target for HOCl in proteins, is comparable to, or greater than, a number of other reactive side chains including Trp, His, and Lys. Though the products of this process have yet to be determined, the rapid oxidation of cystine may be of considerable biological significance as a result of the cleavage of the -S-S- linkage and/or formation of polar products such as RS(dO)SR. Such reactions would be predicted to give rise to major alterations in protein structure, given the key role of disulfide bonds in the maintenance of some protein conformations. With His, both the R-amino and imidazole ring are of approximately equal reactivity, though this is pH dependent, with the reactivity of the imidazole group subject to greater changes than the R-amino function due to the similarity of the imidazole pKa value to the bulk pH at which the reactions were carried out (see Figure 2). Although it has been previously reported that the imidazole ring of the His side chains reacts with HOCl (39, 40), it has not been previously reported that this reaction competes successfully with the oxidation of side chains such as Lys and Trp. Indeed, reaction of histamine with HOCl at pH 8.0 has been reported previously to occur preferentially at the primary (non-imidazole) amino group (40). Considering the rate constants reported in Table 1, it is surprising that this aliphatic amino group competes effectively with the ring nitrogen. Although the pH employed in this previous study (40) would be expected to reduce the reactivity of the imidazole group (cf. Figure 2, and pKa values for the imidazole ring of ∼6.0 and the free amine of 10-11), the extent of reaction at the primary amine, compared to the imidazole ring, would be expected to be e30% at pH 8.0. The discrepancy may arise from the occurrence of chlorine-transfer reactions between amine sites. Such processes are known to occur with some peptides and DNA bases and result in the formation of (thermodynamically more stable) aliphatic amines from the kinetically favored, ring-derived, species (41-43). As the protonated imidazole ring is ca. 10 times less reactive than the neutral form, small variations in the pKa of the imidazole ring with pH will affect the reactivity. The pKa values of protein-bound His residues are known to be dependent on their environ-
Rate Constants for HOCl with Protein Side Chains
ment, and hence the extent of reaction at such residues in proteins may vary markedly. Recent studies have shown that His residues can be important targets for HOCl, as reaction with heme-free myoglobin results in aggregation of the protein (as detected by SDS-PAGE), with this process ameliorated by blocking the imidazole ring of His and the side-chain amine of Lys.2 These data suggest that both His and Lys residues can play a role in HOCl-mediated aggregation. The rate of reaction of the Trp side-chain with HOCl has not been determined previously, though the destruction of this side-chain has often been employed as a marker of HOCl-mediated damage (9, 19, 25, 30, 4448). The absolute rate constants reported here show that oxidation of the Trp side chain is indeed rapid, and precedes chlorination of Tyr [as suggested previously (19)], though reaction is considerably slower than reaction with Met and Cys. Measurement of the oxidation of this residue is therefore unlikely to accurately reflect the extent, or rate, of oxidation of proteins in complex mixtures when only low concentrations of HOCl are present (cf. the depletion of Trp in Figures 5b and 6b). A recent study has reported that reaction of HOCl with apo-A1 results in the loss of Phe, and the formation of chlorinated Phe (27). The kinetic data reported here are inconsistent with direct chlorination of Phe in proteins or peptides, as no reaction of HOCl with 3-phenylpropionic acid was observed over 15 h. The chlorination of Phe observed in this previous study may have occurred during protein hydrolysis, as the acid used in this process (6 M HCl) can react rapidly with chloramines to produce Cl2, which is a potent chlorinating agent for aromatic rings (49). This suggestion is supported by a recent study where hydrolysis of peptides containing Phe residues and Lys chloramines, resulted in regeneration of Lys and the chlorination of Phe (50). The reactivity of HOCl with organic aliphatic amines has been studied previously and the data obtained for these compounds [k 103-104 M-1 s-1 (51)] compares favorably with the rate constants obtained here for the Lys side chain. Furthermore, these values for reaction with R-amino groups agree well with those reported for Gly and Ala (34). The pH dependence of these reactions, and the individual rate constants for the reactions of HOCl and -OCl with the various protonated states of the amino acids, have been determined (34). The pH dependence generally follows a bell-shaped curve, with the maximum rate of reaction occurring at the average of the pKas of the HOCl and amine groups (34). The pH dependence of these reactions was not investigated further in the present study, but a similar trend is believed to hold. Armesto et al. have also undertaken studies with Gly-Gly-OH and Gly-Gly-OEt, which yielded second-order rate constants, k2 ≈ 105 M-1 s-1 (35), consistent with the majority of reaction occurring at the free R-amino group. The kinetic data obtained in this study allows conclusions to be drawn about the likely order, and rate, of consumption of both free amino acids, and protein side chains and backbone sites, on reaction with HOCl. With mixtures of free amino acids, such as those observed in plasma or cell cytosol, free Cys and Met are likely to be consumed most rapidly, via reaction with the side-chain 2 Chapman, A. L., Hawkins, C. L., Davies, M. J., and Kettle, A., unpublished data.
Chem. Res. Toxicol., Vol. 14, No. 10, 2001 1461
thiol and thioether groups, respectively. A rapid rate of consumption of GSH is also predicted, and observed experimentally (20, 52, 53), as a result of reaction with the Cys thiol group. With all other free amino acids and small peptides, the free R-amino group is the major, if not the only, site of reaction. Thus, reaction with the sidechain is essentially irrelevant for free Tyr and Arg, and only occurs to a minor extent with Lys and Trp (where the ratio of side-chain:R-amino reactivity is 1:20 and 1:9, respectively) and His (though this is pH dependent, see above). These data suggests that little, if any, of the 3-ClTyr observed in biological samples (e.g., refs 18 and 32, and see review in ref 54) arises from direct chlorination of the free amino acid. These data also suggest that oxidation of the R-amino group of Tyr to yield HPAA (55, 56) will only occur to any significant extent after complete depletion of free Cys and Met, and proceeds concurrently with depletion of all other free amino acids. The detection of elevated levels of HPAA-modified proteins in early, intermediate, and advanced atherosclerotic lesions (57) suggests first, that aldehydes generated from all other amino acids (58) should also be present, and second, that the flux of HOCl present within such lesions must be large even at very early stages of lesion development. The rate constants for reaction of HOCl with the backbone amides of peptides and proteins vary by 4 orders of magnitude. The most rapid rates are observed for cyclic dipeptides [e.g., cyclo (Gly)2 and cyclo (Ala)2], but incorporation of a charged side chain [e.g., cyclo (Asp)2] decreases the rate by 4-10-fold. This trend is in agreement with data in ref 42, though the absolute values are slightly different. The rate constants for reaction with N-acetyl-blocked linear peptides are smaller than for the cyclic species, though a similar charge effect is observed. The increase in rate constant (per amide bond) on going from blocked mono- to di- to tri-peptides is believed to be due to a greater extent of reaction with midchain groups, which are less deactivated than the termini. The greater reactivity of the cyclic species probably reflects a more favorable conformation for reaction (42). A previous study with N-methylacetamide, the simplest model of a peptide bond, yielded a rate constant, at pH 6.3, of k2 ≈ 0.05 M-1 s-1 (59), which is comparable with the current data. Jensen et al. have also determined, by NMR spectroscopy, a second-order rate constant (k2 1.58 × 10-3 M-1 s-1) for reaction of HOCl with N-acetyl-Ala at pH 7.0 (60), which is in line with the data in Table 2. This group also reported that the chloramide was in equilibrium with the reactants, though the rate constant for the back reaction is very low (k-2 7.57 × 10-7 s-1). Nightingale et al. have investigated the reactions of HOCl with small peptides at pH 7.0 by comparing chlorination of a Lys-containing peptide with other materials (50). This allowed a scale of comparative reactivity, relative to Lys, to be determined. They concluded that the hierarchy of reactivity is Met ≈ Trp > Lys > His ≈ Arg ≈ Tyr, with little backbone cleavage. These observations agree in the main with the current data, with the exception of the preferential reaction of Lys over His. This discrepancy may be due to a neighboring group effect on the pKa of the His residue, leading to a change in reactivity (cf. Figure 2). Chlorine atom transfer to unreacted Lys side chains from initial Hisderived chloramines may also occur in these samples. The kinetic modeling studies reported here for proteins are potentially limited by a number of factors. The first
1462
Chem. Res. Toxicol., Vol. 14, No. 10, 2001
of these arises from the assumption that all sites in a protein are equally accessible to attack. This is unlikely to be completely true, as some residues (e.g., Met and Trp) are often buried preferentially in hydrophobic regions. The second potentially confounding factor arises from the observation that the rate constants, particularly with backbone amides, can be affected by neighboring charged groups. These factors cannot be readily incorporated into this model. Additionally, certain residues, notably His and Cys, have variable pKa values depending on their environment. However, despite these limitations, the results obtained agree well with experimental data. For example, protein fragmentation has been shown experimentally (11) to occur with 70-fold or greater excesses of HOCl over bovine serum albumin (BSA). This figure is in accord with that predicted by the model, where significant reaction at backbone amides, a (presumed) necessary precursor to fragmentation, occurs at around these excesses. Similarly, the predicted order of depletion of residues in HSA is similar to that observed experimentally for both HSA and human plasma (where HSA is the predominant protein) (6, 7). Overall, it is clear that the majority of HOCl consumed by proteins occurs via reaction with reactive side chains and that reaction with backbone amides only occurs to a significant extent with an excess of HOCl over the total number of reactive side-chain residues. The latter is a protein-dependent factor, and therefore, some proteins will be more readily fragmented by HOCl than others; this is in accord with experimental observations (11). Furthermore, as the concentration of reactive side chains on proteins is much greater than the concentrations of free amino acids in plasma or cells (37), the majority of HOCl consumed in most biological situations will be via reaction with proteins. The concordance of the predictions from the modeling and the experimental data suggest that the potentially confounding factors outlined above do not affect the resulting data to a major extent. Furthermore, the percentage of HOCl accounted for by formation of 3-Cl-Tyr residues in proteins (15, 18) is similar to that predicted by the model (ca. 4% with a 170-fold excess of HOCl). The rate constant obtained for the direct chlorination of the Tyr ring, which is the major process that will occur with this amino acid when in a protein due to the slow rate of HOCl with the amide bond nitrogen, is in reasonable agreement with data obtained for model compounds. Thus, reaction of HOCl with (neutral) dihydroxybenzenes (e.g., resorcinol derivatives) are slow (k < 100 M-1 s-1) though reaction with the mono- and dianions are rapid (104-108 M-1 s-1) (61). Rapid reaction is also observed with phenolate anions, with k ca. 105 M-1 s-1 (62, 63). The 3-Cl-Tyr and 3,5-Cl2-Tyr detected in vivo (18, 32) could potentially arise via either direct chlorination of the ring, or chlorine transfer from other species such as chloramines (31, 64). The rate constants in Table 1 suggest that chlorine transfer to the phenolic ring from backbone amides (cf., ref 31) is unlikely to be a major pathway to 3-Cl-Tyr, as reaction of HOCl with backbone amides is slower than direct reaction with the aromatic ring, and only a minor process at most HOCl concentrations. Further studies need to be carried out to determine the kinetics of other potential transfer processes, before the quantitative significance of these two routes is established.
Pattison and Davies
The data obtained in this work allow an estimate to be made of the flux of HOCl (or other chlorinating species) to which proteins are exposed in vivo. Studies on atherosclerotic plaques and abscesses have shown that the conversion of Tyr to 3-Cl-Tyr is 0.04-0.08% (32). The modeling data indicate that a 30-fold molar excess of HOCl over protein would be required for HSA, and a 8-fold molar excess for apo-A1, to obtain this level of Tyr chlorination. If a 20-fold excess is taken as an average value from these two proteins, this would suggest that tissue proteins must be exposed to a total HOCl (or other chlorinating species) exposure of the order of 20 mol/mol over the average lifetime of the proteins [or 50 mM HOCl, if the protein concentration is taken as ca. 2.5 mM, the approximate value for plasma (65)] to generate the observed levels of chlorination. This calculation does not take in to account the possible effect of chlorine transfer to Tyr from other sites, though such reactions would have to be very facile to alter this estimate significantly.
Conclusions The absolute rate constants obtained in this work allow a quantitative evaluation of the role of various potential reactive sites in the reaction of free amino acids, peptides, and proteins with HOCl. The data yield a pattern of reactivity (and their reactive moieties) for proteins of Met (-SMe) > Cys (-SH) . cystine (-SS-) ≈ His (ring) ≈ R-amino > Trp (ring) > Lys (side-chain amino) . Tyr (ring) ≈ Arg (side chain) > backbone amides . Gln ≈ Asn. These kinetic data can be employed to model HOCl reactivity with proteins, given their amino acid composition, and this yields results that are broadly in agreement with experimental results. With free amino acids, a similar scale of reactivity applies for the more reactive side chains, but the less reactive side chains (Tyr, Arg, Lys, Trp) do not compete effectively with reaction at the R-amino group. Furthermore, these data allow estimates to be made as to the flux of HOCl to which proteins are exposed in vivo.
Acknowledgment. We thank Prof. Peter Lay and Dr. Aviva Levina (University of Sydney) for use of the Applied Photophysics SX.18MV stopped-flow system, and Prof. Roger Dean and Dr. Clare Hawkins (Heart Research Institute) for helpful discussions. This work was supported by the Australian Research Council, Grants A00001441 and F00001444.
References (1) Kettle, A. J., and Winterbourn, C. C. (1997) Myeloperoxidase: a key regulator of neutrophil oxidant production. Redox Rep. 3, 3-15. (2) Klebanoff, S. J. (1999) Myeloperoxidase. Proc. Assoc. Am. Phys. 111, 383-389. (3) O’Brien, P. J. (2000) Peroxidases. Chem. Biol. Interact. 129, 113139. (4) Morris, J. C. (1966) The acid ionization constant of HOCl from 5 °C to 35 °C. J. Phys. Chem. 70, 3798-3805. (5) Winterbourn, C. C. (1985) Comparative reactivities of various biological compounds with myeloperoxidase-hydrogen peroxidechloride, and similarity of the oxidant to hypochlorite. Biochim. Biophys. Acta 840, 204-210. (6) Arnhold, J., Hammerschmidt, S., Wagner, M., Mueller, S., Arnold, K., and Grimm, E. (1990) On the action of hypochlorite on human serum albumin. Biomed. Biochim. Acta 49, 991-997. (7) Arnhold, J., Hammerschmidt, S., and Arnold, K. (1991) Role of functional groups of human plasma and luminol in scavenging of NaOCl and neutrophil-derived hypochlorous acid. Biochim. Biophys. Acta 1097, 145-151.
Rate Constants for HOCl with Protein Side Chains (8) Vissers, M. C., and Winterbourn, C. C. (1991) Oxidative damage to fibronectin. I. The effects of the neutrophil myeloperoxidase system and HOCl. Arch. Biochem. Biophys. 285, 53-59. (9) Hazell, L. J., and Stocker, R. (1993) Oxidation of low-density lipoprotein with hypochlorite causes transformation of the lipoprotein into a high-uptake form for macrophages. Biochem. J. 290, 165-172. (10) Hawkins, C. L., and Davies, M. J. (1998) Reaction of HOCl with amino acids and peptides: EPR evidence for rapid rearrangement and fragmentation reactions of nitrogen-centered radicals. J. Chem. Soc., Perkin Trans. 2, 1937-1945. (11) Hawkins, C. L., and Davies, M. J. (1998) Hypochlorite-induced damage to proteins: formation of nitrogen-centred radicals from lysine residues and their role in protein fragmentation. Biochem. J. 332, 617-625. (12) Hawkins, C. L., and Davies, M. J. (1999) Hypochlorite-induced oxidation of proteins in plasma: formation of chloramines and nitrogen-centred radicals and their role in protein fragmentation. Biochem. J. 340, 539-548. (13) Heinecke, J. W., Li, W., Daehnke, H. L., and Goldstein, J. A. (1993) Dityrosine, a specific marker of oxidation, is synthesized by the myeloperoxidase-hydrogen peroxide system of human neutrophils and macrophages. J. Biol. Chem. 268, 4069-4077. (14) Hazen, S. L., Hsu, F. F., Mueller, D. M., Crowley, J. R., and Heinecke, J. W. (1996) Human neutrophils employ chlorine gas as an oxidant during phagocytosis. J. Clin. Invest. 98, 1283-1289. (15) Kettle, A. J. (1996) Neutrophils convert tyrosyl residues in albumin to chlorotyrosine. FEBS Lett. 379, 103-106. (16) Hazen, S. L., d'Avignon, A., Anderson, M. A., Hsu, F. F., and Heinecke, J. W. (1998) Human neutrophils employ the myeloperoxidase-hydrogen peroxide-chloride system to oxidise R-aminoacids to a family of reactive aldehydes. J. Biol. Chem. 273, 49975005. (17) Armesto, X. L., Canle, M., Fernandez, M. I., Garcia, M. V., and Santaballa, J. A. (2000) First steps in the oxidation of sulfurcontaining amino acids by hypohalogenation: very fast generation of intermediate sulfenyl halides and halosulfonium cations. Tetrahedron 56, 1103-1109. (18) Fu, S., Wang, H., Davies, M. J., and Dean, R. T. (2000) Reaction of hypochlorous acid with tyrosine and peptidyl-tyrosyl residues gives dichlorinated and aldehydic products in addition to 3-chlorotyrosine. J. Biol. Chem. 275, 10851-10857. (19) Olszowski, S., Olszowska, E., Stelmaszynska, T., Krawczyk, A., Marcinkiewicz, J., and Baczek, N. (1996) Oxidative modification of ovalbumin. Acta Biochim. Pol. 43, 661-672. (20) Winterbourn, C. C., and Brennan, S. O. (1997) Characterization of the oxidation products of the reaction between reduced glutathione and hypochlorous acid. Biochem. J. 326, 87-92. (21) Harrison, C. A., Raftery, M. J., Walsh, J., Alewood, P., Iismaa, S. E., Thliveris, S., and Geczy, C. L. (1999) Oxidation regulates the inflammatory properties of the murine S100 protein S100A8. J. Biol. Chem. 274, 8561-8569. (22) Yang, C. Y., Gu, Z. W., Yang, M., Lin, S. N., Garcia-Prats, A. J., Rogers, L. K., Welty, S. E., and Smith, C. V. (1999) Selective modification of apoB-100 in the oxidation of low-density lipoproteins by myeloperoxidase in vitro. J. Lipid Res. 40, 686-698. (23) Silverstein, R. M., and Hager, L. P. (1974) The chloroperoxidasecatalyzed oxidation of thiols and disulfides to sulfenyl chlorides. Biochemistry 13, 5069-5073. (24) Davies, M. J., and Hawkins, C. L. (2000) Hypochlorite-induced oxidation of thiols: formation of thiyl radicals and the role of sulfenyl chlorides as intermediates. Free Radical Res. 33, 719729. (25) Naskalski, J. W. (1994) Oxidative modification of protein structures under the action of myeloperoxidase and the hydrogen peroxide and chloride system. Ann. Biol. Clin. 52, 451-456. (26) Yang, C. Y., Gu, Z. W., Yang, H. X., Yang, M., Wiseman, W. S., Rogers, L. K., Welty, S. E., Katta, V., Rohde, M. F., and Smith, C. V. (1997) Oxidation of bovine beta-casein by hypochlorite. Free Radical Biol. Med. 22, 1235-1240. (27) Bergt, C., Oettl, K., Keller, W., Andreae, F., Leis, H. J., Malle, E., and Sattler, W. (2000) Reagent or myeloperoxidase-generated hypochlorite affects discrete regions in lipid-free and lipidassociated human apolipoprotein A-I. Biochem. J. 346, 345-354. (28) Thomas, E. L. (1979) Myeloperoxidase, hydrogen peroxide, chloride antimicrobial system: Nitrogen-Chlorine derivatives of bacterial components in bactericidal action against Escherichia coli. Infect. Immun. 23, 522-531. (29) Thomas, E. L., Grisham, M. B., and Jefferson, M. M. (1986) Preparation and characterization of chloramines. Methods Enzymol. 132, 569-585. (30) Hazell, L. J., van den Berg, J. J. M., and Stocker, R. (1994) Oxidation of low-density lipoprotein by hypochlorite causes ag-
Chem. Res. Toxicol., Vol. 14, No. 10, 2001 1463
(31)
(32)
(33)
(34) (35)
(36) (37)
(38)
(39)
(40)
(41)
(42)
(43)
(44)
(45)
(46)
(47)
(48)
(49) (50)
(51) (52)
(53)
gregation that is mediated by modification of lysine residues rather than lipid oxidation. Biochem. J. 302, 297-304. Domigan, N. M., Charlton, T. S., Duncan, M. W., Winterbourn, C. C., and Kettle, A. J. (1995) Chlorination of tyrosyl residues in peptides by myeloperoxidase and human neutrophils. J. Biol. Chem. 270, 16542-16548. Hazen, S. L., and Heinecke, J. W. (1997) 3-Chlorotyrosine, a specific marker of myeloperoxidase-catalysed oxidation, is markedly elevated in low-density lipoprotein isolated from human atherosclerotic intima. J. Clin. Invest. 99, 2075-2081. Drozdz, R., Naskalski, J. W., and Sznajd, J. (1988) Oxidation of amino acids and peptides in reaction with myeloperoxidase, chloride and hydrogen peroxide. Biochim. Biophys. Acta 957, 4752. Armesto, X. L., Canle, L., and Santaballa, J. A. (1993) R-Amino acids chlorination in aqueous media. Tetrahedron 49, 275-284. Armesto, X. L., Lopez, M. C., Garcia, M. V., Losada, M., and Santaballa, J. A. (1994) Chlorination of dipeptides by hypochlorous acid in aqueous solution. Gazz. Chim. Ital. 124, 519-523. Dawson, R. M. C., Elliott, D. C., Elliott, W. H., and Jones, K. M. (1986) Data for Biochemical Research, Clarendon Press, Oxford. Lentner, C. (1984) Geigy Scientific Tables: Physical Chemistry, Composition of Blood, Hematology, Somatometric Data, Volume 3, Ciba-Geigy, Basle. Peskin, A. V., and Winterbourn, C. C. (2001) Kinetics of the reactions of hypochlorous acid and amino acid chloramines with thiols, methionine, and ascorbate. Free Radical Biol. Med. 30, 572-579. Thomas, E. L., Jefferson, M. M., Bennett, J. J., and Learn, D. B. (1987) Mutagenic activity of chloramines. Mutat. Res. 188, 3543. Thomas, E. L., Jefferson, M. M., Learn, D. B., King, C. C., and Dabbous, M. K. (2000) Myeloperoxidase-catalyzed chlorination of histamine by stimulated neutrophils. Redox Rep. 5, 191-196. Prutz, W. A. (1998) Interactions of hypochlorous acid with pyrimidine nucleotides, and secondary reactions of chlorinated pyrimidines with GSH, NADH, and other substrates. Arch. Biochem. Biophys. 349, 183-191. Prutz, W. A. (1999) Consecutive halogen transfer between various functional groups induced by reaction of hypohalous acids: NADH oxidation by halogenated amide groups. Arch. Biochem. Biophys. 371, 107-114. Hawkins, C. L., and Davies, M. J. (2001) Hypochlorite-induced damage to nucleosides: formation of chloramines and nitrogencentered radicals. Chem. Res. Toxicol. 14, 1071-1081. Schraufstatter, I. U., Browne, K., Harris, A., Hyslop, P. A., Jackson, J. H., Quehenberger, O., and Cochrane, C. G. (1990) Mechanisms of hypochlorite injury of target cells. J. Clin. Invest. 85, 554-562. Jerlich, A., Fabjan, J. S., Tschabuschnig, S., Smirnova, A. V., Horakova, L., Hayn, M., Auer, H., Guttenberger, H., Leis, H. J., Tatzber, F., Waeg, G., and Schaur, R. J. (1998) Human lowdensity lipoprotein as a target of hypochlorite generated by myeloperoxidase. Free Radical Biol. Med. 24, 1139-1148. Hazell, L. J., Davies, M. J., and Stocker, R. (1999) Secondary radicals derived from chloramines of apolipoprotein B-100 contribute to HOCl-induced LDL lipid peroxidation. Biochem. J. 339, 489-495. De Cristofaro, R., and Landolfi, R. (2000) Oxidation of human alpha-thrombin by the myeloperoxidase-H2O2-chloride system: structural and functional effects. Thromb. Haemost. 83, 253-261. Jerlich, A., Hammel, M., Nigon, F., Chapman, M. J., and Schaur, R. J. (2000) Kinetics of tryptophan oxidation in plasma lipoproteins by myeloperoxidase-generated HOCl. Eur. J. Biochem. 267, 4137-4143. March, J. (1992) Advanced Organic Chemistry, John Wiley and Sons, New York. Nightingale, Z. D., Lancha, A. H., Jr., Handelman, S. K., Dolnikowski, G. G., Busse, S. C., Dratz, E. A., Blumberg, J. B., and Handelman, G. J. (2000) Relative reactivity of lysine and other peptide-bound amino acids to oxidation by hypochlorite. Free Radical Biol. Med. 29, 425-433. Antelo, J. M., Arce, F., and Parajo, M. (1995) Kinetic study of the formation of N-chloramines. Int. J. Chem. Kinet. 27, 637-647. Vissers, M. C., and Winterbourn, C. C. (1995) Oxidation of intracellular glutathione after exposure of human red blood cells to hypochlorous acid. Biochem. J. 307, 57-62. Vissers, M. C., Carr, A. C., and Chapman, A. L. (1998) Comparison of human red cell lysis by hypochlorous and hypobromous acids: insights into the mechanism of lysis. Biochem. J. 330, 131-138.
1464
Chem. Res. Toxicol., Vol. 14, No. 10, 2001
(54) Davies, M. J., Fu, S., Wang, H., and Dean, R. T. (1999) Stable markers of oxidant damage to proteins and their application in the study of human disease. Free Radical Biol. Med. 27, 11511163. (55) Hazen, S. L., Hsu, F. F., and Heinecke, J. W. (1996) p-Hydroxyphenylacetaldehyde is the major product of L-tyrosine oxidation by activated human phagocytes. A chloride-dependent mechanism for the conversion of free amino acids into reactive aldehydes by myeloperoxidase. J. Biol. Chem. 271, 1861-1867. (56) Hazen, S. L., Gaut, J. P., Hsu, F. F., Crowley, J. R., d’Avignon, A., and Heinecke, J. W. (1997) p-Hydroxyphenylacetaldehyde, the major product of L-tyrosine oxidation by the myeloperoxidaseH2O2-chloride system of phagocytes, covalently modifies epsilonamino groups of protein lysine residues. J. Biol. Chem. 272, 16990-16998. (57) Hazen, S. L., Gaut, J. P., Crowley, J. R., Hsu, F. F., and Heinecke, J. W. (2000) Elevated levels of protein-bound p-hydroxyphenylacetaldehyde, an amino- acid-derived aldehyde generated by myeloperoxidase, are present in human fatty streaks, intermediate lesions and advanced atherosclerotic lesions. Biochem. J. 352, 693-699. (58) Hazen, S. L., Hsu, F. F., d’Avignon, A., and Heinecke, J. W. (1998) Human neutrophils employ myeloperoxidase to convert alphaamino acids to a battery of reactive aldehydes: a pathway for
Pattison and Davies
(59) (60) (61) (62) (63)
(64) (65)
aldehyde generation at sites of inflammation. Biochemistry 37, 6864-6873. Antelo, J. M., Arce, F., Parajo, M., Pousa, A. I., and Perez-Moure, J. C. (1995) Chlorination of N-methylacetamide: a kinetic study. Int. J. Chem. Kinet. 27, 1021-1031. Jensen, J. S., Lam, Y.-F., and Helz, G. R. (1999) Role of amide nitrogen in water chlorination: proton NMR evidence. Environ. Sci. Technol. 33, 3568-3573. Rebenne, L. M., Gonzalez, A. C., and Olson, T. M. (1996) Aqueous chlorination kinetics and mechanism of substituted dihydroxybenzenes. Environ. Sci. Technol. 30, 2235-2242. Soper, F. G., and Smith, G. F. (1926) The halogenation of phenols. J. Chem. Soc., 1582-1591. Lee, F. C. (1967) Kinetics of reactions between chlorine and phenolic compounds. In Principles and Applications of Water Chemistry (Faust, S. D., and Hunter, J. V., Eds.) pp 54-74, Wiley, New York. Haberfield, P., and Paul, D. (1965) The chlorination of anilines. Proof of the existence of an N-chloro intermediate. J. Am. Chem. Soc. 87, 5502. Vander, A. J., Sherman, J. H., and Luciano, D. S. (1980) Human Physiology, McGraw-Hill, New York.
TX0155451