Preventing Protein Oxidation with Sugars: Scavenging of Hypohalous

Oct 17, 2012 - The Heart Research Institute, 7 Eliza Street, Newtown, NSW 2042, Australia ...... Vaisar , T., Shao , B., Green , P. S., Oda , M. N., O...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/crt

Preventing Protein Oxidation with Sugars: Scavenging of Hypohalous Acids by 5‑Selenopyranose and 4‑Selenofuranose Derivatives Corin Storkey,†,∥ David I. Pattison,†,‡,∥ Jonathan M. White,§ Carl H. Schiesser,§,∥ and Michael J. Davies*,†,‡,∥ †

The Heart Research Institute, 7 Eliza Street, Newtown, NSW 2042, Australia Faculty of Medicine, University of Sydney, Sydney, NSW 2006, Australia § School of Chemistry and Bio21 Molecular Science and Biotechnology Institute, The University of Melbourne, Victoria 3010, Australia ∥ ARC Centre of Excellence for Free Radical Chemistry and Biotechnology, Australia ‡

S Supporting Information *

ABSTRACT: Heme peroxidases including myeloperoxidase (MPO) are released at sites of inflammation by activated leukocytes. MPO generates hypohalous acids (HOX, X = Cl, Br, SCN) from H2O2; these oxidants are bactericidal and are key components of the inflammatory response. However, excessive, misplaced or mistimed production can result in host tissue damage, with this implicated in multiple inflammatory diseases. We report here methods for the conversion of simple monosaccharide sugars into selenium- and sulfur-containing species that may act as potent water-soluble scavengers of HOX. Competition kinetic studies show that the seleno species react with HOCl with rate constants in the range 0.8− 1.0 × 108 M−1 s−1, only marginally slower than those for the most susceptible biological targets including the endogenous antioxidant, glutathione. The rate constants for the corresponding sulfur-sugars are considerably slower (1.4−1.9 × 106 M−1 s−1). Rate constants for reaction of the seleno-sugars with HOBr are ∼8 times lower than those for HOCl (1.0−1.5 × 107 M−1 s−1). These values show little variation with differing sugar structures. Reaction with HOSCN is slower (∼102 M−1 s−1). The selenosugars decreased the extent of HOCl-mediated oxidation of Met, His, Trp, Lys, and Tyr residues, and 3-chlorotyrosine formation, on both isolated bovine serum albumin and human plasma proteins, at concentrations as low as 50 μM. These studies demonstrate that novel selenium (and to a lesser extent, sulfur) derivatives of monosaccharides could be potent modulators of peroxidase-mediated damage at sites of acute and chronic inflammation, and in multiple human pathologies.

1. INTRODUCTION

phagocytic cells that eliminate parasites and related organisms. Unlike neutrophils, which phagocytose their target organisms and subsequently release MPO primarily into the phagolysosomal compartment, eosinophils exocytose their granule contents on to the parasite surface to which they are attached.3 Salivary peroxidase or LPO is found in multiple human exocrine secretions including tears, milk, saliva, and vaginal fluid. In each case, the primary role of these peroxidases is as a first line of defense against invading microorganisms.4 For each peroxidase, the proportions of HOCl/HOBr/ HOSCN produced are determined by the selectivity constants of the enzymes for each substrate anion, and the physiological concentrations of Cl−, Br−, and SCN−. At neutral pH and normal physiological plasma anion concentrations, ∼45% of the

There is considerable interest in the roles played by mammalian heme peroxidase enzymes (e.g., myeloperoxidase (MPO), eosinophil peroxidase (EPO), and lactoperoxidase (LPO)), in both the innate immune system, where these species play a critical role in killing invading pathogens, and in human pathologies associated with chronic inflammation. MPO is released at sites of inflammation from intracellular granules by activated neutrophils, monocytes, and some tissue macrophages.1 Activation of these cells results in the generation of hydrogen peroxide (H2O2) by NADPH oxidase enzymes via a respiratory burst.2 MPO utilizes this H2O2 to oxidize halide and pseudohalide ions, predominantly chloride (Cl−), bromide (Br−), and thiocyanate (SCN−), to generate the potent oxidants, hypochlorous (HOCl), hypobromous (HOBr), and hypothiocyanous acids (HOSCN), respectively. EPO is a major granule protein of eosinophils, which are specialized human © 2012 American Chemical Society

Received: August 17, 2012 Published: October 17, 2012 2589

dx.doi.org/10.1021/tx3003593 | Chem. Res. Toxicol. 2012, 25, 2589−2599

Chemical Research in Toxicology

Article

seleno-L-talitol [7 (SeTal)] from glucose and mannose, respectively, are described, together with a kinetic and product analysis evaluation of these compounds as water-soluble protective agents against HOCl-, HOBr-, and HOSCNmediated protein damage.

H2O2 consumed by MPO is used to generate HOCl, 50% HOSCN, and the remaining 5% HOBr.5 In contrast, EPO generates predominantly HOBr and HOSCN, while LPO favors HOSCN.6,7 These oxidants are bactericidal or bacteriostatic and are key components of the inflammatory response. These properties underlie the widespread use, particularly of HOCl (a major active component of household bleach), in the water purification, hygiene, and disinfection fields.8,9 Excessive or misplaced production of these species in vivo has, however, been linked to several human pathologies as a result of collateral damage to host tissues.10 The role of MPO in the pathogenesis of disease has been researched extensively, and considerable evidence links this enzyme and the oxidants that it generates with the pathogenesis of cardiovascular disease (atherosclerosis), cystic fibrosis, sepsis, rheumatoid arthritis, asthma, kidney disease, and some cancers.11−16 The evidence for a role for MPO in the pathogenesis of atherosclerosis is compelling,11,17,18 with clinical studies indicating that elevated plasma and blood MPO levels are a strong independent risk factor and predictor of outcomes for cardiovascular disease.12 Furthermore, it has been reported that MPO mRNA, protein, and peroxidase activity are present in atherosclerotic lesions, that elevated levels of chlorinated tyrosine residues (3-chloro-Tyr, an established biomarker of HOCl damage) are present on proteins extracted from diseased human arteries compared to healthy tissue, and that there is a direct correlation between the extent of HOCl-damaged proteins and disease severity.19−22 HOCl and HOBr are powerful oxidants (two-electron reduction potentials of 1.28 and 1.13 V, respectively)23 that react rapidly with biological molecules, particularly with sulfurcenters and amine functions on proteins, unsaturated lipids, antioxidants, and DNA.24 In contrast, HOSCN is a less powerful oxidant (two-electron reduction potential of 0.56 V)23 that reacts almost exclusively with sulfur and selenoresidues.25−29 Kinetic data are consistent with proteins being major targets for HOCl within biological systems as a result of their abundance and high reactivity.30 Exposure of isolated proteins to HOCl results predominantly in modification at sidechains (Cys, Met, cystine, His, Lys, Trp, and Tyr), with limited protein fragmentation and aggregation observed at high oxidant concentrations.10 We have recently demonstrated that the carbohydrate derivatives 1,5-dideoxy-5-thio/seleno-L-gulitol [1 (SGul), 2 (SeGul)] and 1,5-dideoxy-5-thio/seleno- D -mannitol [3 (SMan), 4 (SeMan)] (Figure 1) afford significant protection to human plasma proteins against HOCl in vitro.31 Here, the syntheses of a range of sulfur- and selenium-containing carbohydrate derivatives, 1,5-dideoxy-5-thio/seleno-L-iditol [5 (SId), 6 (SeId)] and the furanose derivative 1,4-dideoxy-4-

2. EXPERIMENTAL PROCEDURES 2.1. Synthesis and Characterization of Novel Compounds. The strategies used to synthesize compounds 5−7 are outlined in detail in the Results section. NMR, IR, elemental analysis, mass spectrometry, optical rotation for all compounds (5−7, 9−12, 14− 15), and X-ray crystal structure data for 7 are given in Supporting Information. The data obtained for compounds 9, 10, 14, and 15 agree with literature reports.32−34 1 H NMR spectra were recorded on Varian Inova 400 (400 MHz) or Varian Inova 500 (500 MHz) instruments at room temperature, using CDCl3 or CD3OD as an internal reference deuterium lock, CDCl3 at δ 7.26 ppm and CD3OD at δ 3.31 ppm. The chemical shift data for each signal are given as δ in units of parts per million (ppm). The multiplicity of each signal is indicated by s (singlet), d (doublet), t (triplet), q (quartet), dd (doublet of doublets), dt (doublet of triplets), and m (multiplet). The number of protons (n) for a given resonance is indicated by n H. Coupling constants (J) are quoted in Hz and are recorded to the nearest 0.1 Hz. 13C NMR spectra were recorded on Varian Inova 400 (400 MHz) or Varian Inova 500 (500 MHz) instruments using the central resonance of the triplet of CDCl3 at δ 77.23 ppm or the septet of CD3OD at δ 49.00 as an internal reference. The chemical shift data for each signal are given as δ in units of parts per million (ppm). Infrared spectra were recorded on a Perkin-Elmer Spectrum One FT-IR spectrometer in the region 4000−650 cm−1. The samples were analyzed as thin films from dichloromethane or methanol. Mass spectra were recorded at the Bio21 Institute, The University of Melbourne. Low-resolution spectra were recorded on a Waters Micromass Quattro II instrument (EI and CI). High-resolution mass spectrometry experiments were conducted using a hybrid linear ion trap and Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometer (Finnigan LTQ-FT San Jose, CA), equipped with ESI. Ions of interest were mass selected in the LTQ using standard procedures and analyzed in the FT-ICR MS to generate the highresolution tandem mass spectrum. Optical specific rotations were measured using a Jasco DIP-1000 digital polarimeter, in a cell of 1 cm path length. The concentration (c) is expressed in g/100 cm3 (equivalent to g/0.1 dm3). Specific rotations are denoted [α] and given in implied units of °dm2·g−1 (T = temperature in °C). Analytical thin layer chromatography (TLC) was carried out on precoated 0.25 mm thick Merck 60 F254 silica gel plates. Visualization was by absorption of UV light or thermal development after dipping in an ethanolic solution of phosphomolybdic acid (PMA) or sulfuric acid (H2SO4). Flash chromatography was carried out on silica gel [Merck Kieselgel 60 (230−400 mesh)] under nitrogen pressure. Hydrogenation was carried out in a Büchi GlasUster “miniclave drive” stainless steel vessel, 100 mL, with a maximum operation pressure of 60 bar. Teflon inserts were used, and reactions were stirred using magnetic stirrer bars. Crystals of compound 7 were mounted in low temperature oil and flash cooled to 130 K using an Oxford Cryostream low temperature device. Intensity data were collected at 130 K on an Oxford SuperNova X-ray diffractometer with a CCD detector using Cu−Kα radiation (λ = 1.54184 Å). Data were reduced and corrected for absorption (CrysAlis CCD, Oxford Diffraction Ltd., version 1.171.32.5 (release 08-05-2007 CrysAlis171 .NET)). The structures were solved by direct methods and difference Fourier synthesis using the SHELX suite of programs,35 as implemented within the WINGX5 software.36 Thermal ellipsoid plots were generated using the program ORTEP-3. 2.2. Biochemical Reagents. All chemicals were obtained from Sigma/Aldrich/Fluka and were used as received, with the exception of sodium hypochlorite (in 0.1 M NaOH, low in bromine; BDH Chemicals). HOCl was quantified by its absorbance at 292 nm at pH 12 using ε292 = 350 M−1 cm−1.37 All studies were performed in 10 mM

Figure 1. Sulfur- and seleno-sugar derivatives prepared and investigated in these studies. 2590

dx.doi.org/10.1021/tx3003593 | Chem. Res. Toxicol. 2012, 25, 2589−2599

Chemical Research in Toxicology

Article

phosphate buffer (pH 7.4). Phosphate buffers were prepared using Milli-Q water treated with washed Chelex resin (Bio-Rad) to remove contaminating transition metal ions. The pH values of solutions were adjusted, where necessary, to pH 7.4 using 100 mM H2SO4 or 100 mM NaOH. 2.3. Competitive Kinetic Studies for Thio/Seleno-Sugars against HOCl Using FMoc-Met. The kinetics of the reactions of HOCl (1 μM) with the thio/seleno-sugar derivatives (“scavenger”; 1.2 μM−20 μM) were investigated in competition with Fmoc-Met (5 μM) at 22 °C using an adaptation of a previous method,38 in which the conversion of Fmoc-Met to the Fmoc-Met sulfoxide is quantified, in the absence and presence of a putative scavenger. The yields of FmocMet sulfoxide generated in the presence of increasing concentrations of the sugar derivative (yieldscavenger) were determined by UPLC and compared to the maximal yield in the absence of added sugar/ scavenger (yieldmax). Using a competition kinetic analysis, the yields are related by eq 1, and rearrangement of this equation results in the linear form (y = mx + c), given in eq 2.

yield scavenger yield max − yield scavenger yield max [FmocMet] yield scavenger

=

=

kFmocMet[FmocMet] kscavenger[scavenger]

kscavenger[scavenger] kFmocMet

size; Rockland Technologies, Newport, DE) packed with octadecyl silanized silica, equipped with a Pelliguard guard column (2 cm; Supelco), at 30 °C (using a column oven, Waters Corp., Milford, MA), with a solvent flow rate of 1 mL·min−1. The mobile phase comprised a gradient of solvent A (10 mM phosphoric acid with 100 mM sodium perchlorate, pH 2.0) and solvent B [80% (v/v) MeOH in nanopure water], programmed as follows: 20% solvent B and 80% solvent A at 0 min increasing to 80% solvent B over 10 min; over the next 5 min, the proportion of solvent B was held at 80%, before the proportion of solvent B was reduced to 20% over 4 min, and the column was allowed to re-equilibrate for 6 min prior to injection of the next sample. Samples were filtered (0.2 μm) to remove particulate matter before analysis, with 50 μL injected for each run. The eluent was monitored using a UV detector (λ 280 nm) and an electrochemical detector (Antec Leyden Intro). The electrochemical detector was set to +1200 mV to quantify the halogenated N-Ac-Tyr products. Peak areas were quantified using Class VP 7.4 SP1 software (Shimadzu) and compared to authentic standards. Under these conditions, N-Ac-Tyr was detected in the UV and electrochemical channel at 8.3 min, N-Ac-3Br-Tyr at 11.4 min, and N-Ac-3,5-diBr-Tyr at 13.4 min. 2.7. Competitive Kinetic Studies for SeTal (7) against HOSCN Using 5-Thio-2-nitrobenzoic Acid (TNB). HOSCN was prepared by in situ LPO-catalyzed reaction of H2O2 with SCN−.28,39 Thus, LPO (1.5−2 μM) was incubated with H2O2 (3.75 mM) and NaSCN (7.5 mM) in 10 mM potassium phosphate buffer (pH 6.6) for 15 min at 22 °C. Catalase (1 mg·mL−1; from bovine liver) was subsequently added to remove unreacted H2O2, followed by centrifugation at 11300g for 5 min through a 10 kDa molecularmass cutoff filter to remove catalase and LPO. HOSCN concentrations were determined by oxidation of TNB (5-thio-2-nitrobenzoic acid) at 412 nm using ε = 14150 M−1·cm−1.40,41 2.8. Kinetic Measurements of HOSCN Reactions. Stopped-flow studies were carried out using an Applied Photophysics SX.18MV stopped-flow system as described previously.42−44 HOSCN was kept as the limiting reagent with a minimum of a 4-fold excess of the substrate. The sample chamber was maintained at 22 °C, with ten runs averaged to improve the signal-to-noise ratio. The rate constant for the reaction of HOSCN with SeTal (7) was determined by competition kinetics against TNB at 412 nm using kTNB = 3.8 × 105 M−1·s−1.25 2.9. HPLC Amino Acid Analysis of HOCl-Oxidized BSA and Plasma. Amino acid analyses of HOCl oxidized BSA and plasma were carried out essentially as described previously.45 BSA (0.5 mg mL−1, ∼7.6 μM) or human plasma diluted to give a final plasma protein concentration of 0.5 mg mL−1 (ca. 7.6 μM, chosen to allow comparison with the isolated protein data) was mixed with buffer (labeled 0 mM) or increasing concentrations of the Se-sugars (2, 4, 6, or 7; 50 μM−1 mM), before reaction with a 100-fold molar excess of HOCl (760 μM), or buffer (labeled “control”). Samples, were incubated for 3 h at 37 °C before being delipidated and the proteins precipitated by the addition of 25 μL 0.3% (w/v) deoxycholic acid and 50 μL of 50% (w/v) TCA. The protein pellets were washed once with 5% (w/v) TCA and twice with ice cold acetone, before being resuspended in 150 μL of 4 M methanesulfonic acid (MSA) containing 0.2% w/v tryptamine. The samples were then hydrolyzed under vacuum at 110 °C for 16−18 h. After neutralization with 150 μL of 4 M NaOH and filtration (centrifuged at 11300g for 2 min through a PVDF 0.22 μm membrane; 0.5 mL volume; No. UFC30GVNB, Millipore), the samples were diluted into water (10-fold) and 40 μL transferred to HPLC vials for analysis. OPA reagent (Sigma-Aldrich, P7914) was activated immediately before use by the addition of 5 μL of 2-mercaptoethanol to 1 mL of OPA reagent in an HPLC vial. Five micromolar standards were prepared by the addition of 10 μL of Sigma-Aldrich amino acid standards (A9781, 500 μM stock) to 990 μL of water. From these stock solutions, 1, 2, 3, 4, and 5 μM standards were prepared for comparative analysis. Analysis and quantification of protein reaction products were carried out on a Shimadzu Nexera UPLC system (Shimadzu, South Rydalmere, NSW, Australia). The reaction mixtures were separated on a Shim-pack XR-ODS (Shimadzu, 100 × 4.6 mm, 2.2 μm) column,

(1)

+ [FmocMet] (2)

From a plot of [Fmoc-Met]·yieldmax/yieldscavenger against increasing concentrations of the scavenger ([scavenger]), the slope of the corresponding line allows determination of the value of kscavenger (i.e., k for reaction of HOCl with the sugar derivative) using kFmocMet (1.3 × 108 M−1 s−1) with a set y-intercept equal to [Fmoc-Met]. 2.4. UPLC Instrumentation and Methods. Quantification of Fmoc-Met and Fmoc-Met sulfoxide were carried out on a Shimadzu Nexera UPLC system (Shimadzu, South Rydalmere, NSW, Australia), with samples separated on a Shim-pack XR-ODS (Shimadzu, 100 × 4.6 mm, 2.2 μm) column at 40 °C with a solvent flow rate of 1.2 mL·min−1. The mobile phase comprised of a gradient of solvent A [(MeOH (20%), THF (2.5%), 1 M NaOAc, pH 5.3 (5%) and H2O (72.5%)] and solvent B [MeOH (80%), THF (2.5%), 1 M NaOAc, pH 5.3 (5%), H2O (12.5%)], programmed as follows: 75% solvent B and 25% solvent A at 0 min, increasing to 87.5% solvent B over 5 min, followed by a further increase to 100% solvent B over the next 0.5 min and a wash with 100% solvent B for 2.5 min, before returning to 75% solvent B over the next 0.5 min with 3.5 min of re-equilibrating preceding the next injection. Samples were filtered (0.2 μm) to remove particulate matter before analysis with 5 μL injected for each run. The eluent was monitored by fluorescence detection (RF-20AXS; λex 265 nm; λem 310 nm), with peak areas determined using Lab Solutions 5.32 SP1 software (Shimadzu) and compared to authentic standards as required. Using these conditions, Fmoc-Met sulfoxide was detected at a retention time of 1.7 min, and Fmoc-Met at 2.8 min. 2.5. Competitive Kinetic Studies for Seleno-Sugars Against HOBr Using N-AcTyr. HOBr was prepared by mixing HOCl (40 mM in water, pH 13) with NaBr (45 mM in water) in equal volumes. The reaction was left for 1 min before dilution to the required concentration of HOBr (typically 0.1 − 0.2 mM). The kinetics of the reactions of HOBr (10 μM) with the thio/seleno-sugar derivatives (20−150 μM) were investigated in competition with N-Ac-Tyr (1 μM) at 22 °C by adapting a previous method,38 using the conversion of N-Ac-Tyr to N-Ac-3-Br-Tyr, in the absence and presence of a scavenger. The yields of N-Ac-3-Br-Tyr at increasing carbohydrate derivative concentration (yieldscavenger) were determined by HPLC and compared to the maximal yield in the absence of added scavenger (yieldmax). The data were subsequently analyzed in an analogous manner to that described for HOCl/Fmoc-Met above, with kN‑Ac‑Tyr = 2.6 × 105 M−1 s−1.38 2.6. HPLC Instrumentation and Methods. Quantification of NAc-Tyr and N-Ac-3-Br-Tyr were carried out on a Shimadzu LC-10A HPLC system (Shimadzu, South Rydalmere, NSW, Australia), on a Zorbax reverse-phase HPLC column (25 cm × 4.6 mm, 5 μm particle 2591

dx.doi.org/10.1021/tx3003593 | Chem. Res. Toxicol. 2012, 25, 2589−2599

Chemical Research in Toxicology

Article

Scheme 1. Reagents and Conditions for the Chemical Synthesis of 1,5-Dideoxy-5-thio/seleno-L-iditol (5 and 6)a

a (i) BnOH, AcCl; (ii) p-TSA, 2-methoxypropene, Me2CO; (iii) H2, Pd/C, EtOH; (iv) NaBH4, EtOH; (v) MsCl, DMAP, CH2Cl2/Py; (vi) Na2S, DMF (11) or NaBH4, Se, EtOH (12); (vii) TFA, CH2Cl2.

Scheme 2. Reagents and Conditions for the Chemical Synthesis of 1,4-Dideoxy-4-seleno-D-talitol (7)a

a

(i) p-TSA, 2,2-dimethoxypropane, Me2CO; (ii) NaBH4, MeOH; (iii) MsCl, DMAP, CH2Cl2/Py; (iv) NaBH4, Se, EtOH; (v) TFA, CH2Cl2.

maintained at 40 °C with a flow rate of 1.2 mL·min−1. The mobile phase comprised a gradient of solvent A [(MeOH (20%), THF (2.5%), 1 M NaOAc, pH 5.3 (5%), and H2O (72.5%)] and solvent B [MeOH (80%), THF (2.5%), 1 M NaOAc, pH 5.3 (5%), and H2O (12.5%)] programmed as follows: 0% solvent B and 100% solvent A at 0 min, increasing to 25% solvent B over 6 min, maintaining 25% for a further 1 min, followed by an increase to 62% solvent B over the next 0.5 min and holding this concentration for another 2.5 min, followed by an increase to 100% B over 2 min, holding at 100% B for another 1 min, before returning to 0% solvent B over the next 0.5 min, and reequilibrating at 0% for 3.5 min. The autoinjector was programmed to add 20 μL of activated OPA reagent to the specified sample (40 μL), followed by 3 mixing cycles, and a 1 min incubation period. After the incubation step, 15 μL of the final reaction mixture was injected. The fluorescence detector was set with λex of 340 nm and λem of 440 nm to detect the OPA tag. The concentration of each amino acid in the samples was determined from linear plots of the HPLC peak area

versus concentration from the standards. Any variation in the efficiency of hydrolysis or sample recovery after the precipitation and washing steps was taken into account by expressing the concentration of the amino acids of interest as a ratio with isoleucine (Ile), as it is not modified by HOCl. 2.10. Analysis of 3-Chlorotyrosine (3-Cl-Tyr) Using LCMS. Sample preparation for 3-Cl-Tyr analysis was identical to that used for total amino acid analysis (see above), except that hydrolysis was achieved by the addition of 150 μL of 6 M HCl and 50 μL of thioglycolic acid into the PicoTag vessel, before being placed under vacuum in the oven at 110 °C for 16−18 h. The hydrolyzed samples were then placed in 1.5 mL centrifuge tubes and dried under vacuum. Each sample was resuspended in 50 μL of water and filtered (centrifuged at 11300g for 2 min through a PVDF 0.22 μm membrane, 0.5 mL volume, No. UFC30GVNB, Millipore) to remove any insoluble precipitate. The samples were then transferred to HPLC vials for LC-MS analysis. 2592

dx.doi.org/10.1021/tx3003593 | Chem. Res. Toxicol. 2012, 25, 2589−2599

Chemical Research in Toxicology

Article

Standard solutions were prepared in Milli-Q water to give 2.5 mM Tyr and 100 μM 3-Cl-Tyr. The two stock solutions were combined and diluted to give 1:1 mixtures of tyrosine: 3-Cl-Tyr with concentrations of 100−500 pmol in 20 μL. Then, 40 μL of each standard was transferred to HPLC vials for LC-MS analysis. Tyr and 3Cl-Tyr were quantified by LC-MS in the positive ion mode with a Thermo LCQ Deca XP ion-trap instrument coupled to a Finnigan Surveyor HPLC system as described previously.46

(pH 7.4). These conditions were chosen to enable direct comparison with previous kinetic data for other targets including Met, Cys, and glutathione (GSH).24,30 Reactions with selenomethionine (SeMet), a well-established oxidant scavenger,47 were also investigated to assess the influence of the ring structure on the reactivity of the selenoether moiety. Reactions of HOCl (1 μM) with the S-/Se-sugar derivatives (1−7; 1.2 μM−100 μM) or SeMet (1.2 μM−20 μM) were investigated using an adaptation of previously established competition kinetics approach, with Fmoc-Met (5 μM) as the competing target.38 A decrease in Fmoc-Met sulfoxide was detected with increasing concentrations of (1−7) (yieldscavenger) with these data compared to the maximal yield in the absence of added scavenger (yieldmax). The linear graphs (Figure 2a) obtained by plotting [Fmoc-Met]·(yieldmax/yieldscavenger) against the sugar concentration ([scavenger]) gave a slope from which the rate constant for reaction of HOCl with the S-/Se-sugar derivatives (Table 1) could be calculated, using k(HOCl+Fmoc‑Met)

3. RESULTS 3.1. Synthesis. Compounds 1−4 were prepared as reported previously.31 The synthesis of 1,5-dideoxy-5-thio/seleno-L-iditol (5 and 6) followed an analogous synthetic route except with the inclusion of an anomeric benzyl protecting group. Accordingly, D-glucose (8) was reacted with benzyl alcohol in the presence of acetyl chloride, followed by further reaction of the isolated crude product with 2-methoxypropene in the presence of catalytic p-toluenesulfonic acid under anhydrous conditions to afford the 1-benzyl-2,3,4,6-di-O-isopropylidene-Dglucoside (9) as the major product in 80% yield over two steps (Scheme 1). The benzyl group was then cleaved by hydrogenation before the crude reaction mixture was reacted with sodium borohydride in ethanol to yield the diol (10), which was purified by chromatography in an overall yield of 90% over two steps. The subsequent step involved conversion of the diol (10) to the corresponding bis-mesylate intermediate (10a), by reaction with methanesulfonyl chloride in dichloromethane and pyridine with catalytic 4-dimethylaminopyridine (DMAP). The bis-mesylate (10a) was then reacted with either sodium sulfide in DMF at 100 °C to give the protected S-sugar (11) in a 74% yield or with selenium metal and sodium borohydride in EtOH at 70 °C to give the protected Se-sugar (12) in 53% yield. Deprotection of 11 and 12 was achieved by treatment with trifluoroacetic acid in dichloromethane, and the resulting residues were purified by flash chromatography to yield SId (5) and SeId (6) as colorless oils. Characterization data for these materials are provided in Supporting Information. The synthesis of the pyranose 1,4-dideoxy-4-seleno-D-talitol (7) began from D-mannose (13). This was reacted with 2,2dimethoxypropane in the presence of catalytic p-toluenesulfonic acid and acetone before reduction of the hemiacetal with sodium borohydride in methanol to yield the diol (14) as the major product in 82% over two steps (Scheme 2). The diol (14) was subsequently converted to the corresponding bismethanesulfonate intermediate (14a), by reaction with methanesulfonyl chloride in dichloromethane and pyridine with catalytic 4-dimethylaminopyridine (DMAP). Reaction of this bis-methanesulfonate species (14a) with selenium and sodium borohydride in EtOH at 70 °C gave the isopropylideneprotected species (15) in 55% yield. Deprotection of (15) was achieved by treatment with trifluoroacetic acid in dichloromethane, and the resulting residue was purified by flash chromatography to yield the desired seleno-sugar derivative SeTal (7) as a colorless oil. Full characterization data for (7) (NMR, IR, elemental analysis, mass spectrometry, optical rotation, and X-ray crystal structure) are provided in Supporting Information. 3.2. Kinetic Studies with HOCl, HOBr, and HOSCN. The reactivity of the thio- and seleno-containing pyranose carbohydrate derivatives (1−7) with HOCl, HOBr, and HOSCN was investigated and the second-order rate constants for their reactions determined at 22 °C and physiological pH

Figure 2. Kinetic data for the reaction of S-/Se-sugar derivatives 1, 2, and 7 with (a) HOCl and (b) HOBr. Representative data for three sugar (“scavenger”) derivatives, SGul (1; ■), SeGul (2; ▲), and SeTal (7; ●), are presented. (a) HOCl scavenging was measured by competition against Fmoc-Met using UPLC with fluorescence detection to quantify Fmoc-Met sulfoxide yields. Slope = k(HOCl + sugar)/k(HOCl + Fmoc-Met) with k(HOCl + Fmoc-Met) = 1.3 × 108 M−1 s−1.19 The insert shows expanded data for SeGul (2) and SeTal (7). (b) HOBr scavenging was measured in competition with N-AcTyr using HPLC with electrochemical detection to quantify N-AcBrTyr formation. Slope = k(HOBr + sugar)/k(HOBr + N-Ac-Tyr) with k(HOBr + N-Ac-Tyr) = 2.6 × 105 M−1 s−1.18 Errors bars for all points are expressed as standard error of the mean (n > 4); in some cases, the error bars are smaller than the symbols. 2593

dx.doi.org/10.1021/tx3003593 | Chem. Res. Toxicol. 2012, 25, 2589−2599

Chemical Research in Toxicology

Article

the small changes in absorbance, an accurate value for the second-order rate constant could not be determined, but an estimate was obtained using eq 3 (with kTNB 3.8 × 105 M−1 s−1)25 for each set of concentrations. It has also been noted previously that analysis of rate constants of HOSCN using TNB as a competitive substrate is complicated by rapid secondary reactions of the TNB sulfenyl thiocyanate (or sulfenic acid) product with a further TNB molecule to form the disulfide, DTNB.26 With these limitations in mind, an approximate second-order rate constant was determined for reaction of HOSCN with 7 of ∼102 M−1 s−1 (Table 1). As this rate constant could not be assessed accurately, experiments with 1−6 were not undertaken.

Table 1. Summary of the Second Order Rate Constants for Scavenging of HOCl, HOBr, and HOSCN by the S-/SeSugar Derivatives (1−7), SeMet, Met, Cys, and GSHa compd SeGul 2 SeMan 4 SeId 6 SeTal 7 SeMet Met SGul 1 SMan 3 SId 5 Cys GSH

k (HOCl)/107 M−1 k (HOBr)/107 M−1 s−1 s−1 9.4 (±0.3) 8.0 (±0.5) 9.0 (±0.3) 10 (±2) 32 (±1) 2.95 (±0.07)c 0.190 (±0.007) 0.141 (±0.006) 0.140 (±0.002) 31.2 (±0.6)c 10.9 (±0.1)c

1.3 (±0.2) 1.09 (±0.03) 1.33 (±0.03) 1.52 (±0.06) 1.4 (±0.1) 0.36 (±0.03)d 0.112 (±0.004) 0.099 (±0.003) 0.115 (±0.004) 1.2 (±0.2)f

k (HOSCN)/104 M−1 s−1

∼0.01 0.28 (±0.02)b