Environ. Sci. Technol. 2007, 41, 2338-2342
Stability of Metal-Glutathione Complexes during Oxidation by Hydrogen Peroxide and Cu(II)-Catalysis
Downloaded via UNIV OF LOUISIANA AT LAFAYETTE on November 5, 2018 at 05:40:20 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
HEILEEN HSU-KIM* Department of Civil & Environmental Engineering, Duke University, Durham, North Carolina 27708
Thiol-containing ligands such as glutathione (GSH) are expected to degrade in the presence of oxygen; however, complexation by Hg, Ag, and other trace metals may protect free thiol functional groups (R-S-) from oxidation, leading to persistence in surface water environments. In this study, the stability of GSH complexes with Hg2+, Ag+, and other metals including Cd2+, Zn2+, and Pb2+ was assessed during exposure to two potential environmental oxidants: H2O2 and Cu2+. The results indicated that Hg(GSH)2 and Ag(GSH) complexes were completely stable for at least 2 days in the presence of either H2O2 or Cu2+. In contrast, free GSH oxidized within minutes to hours. Complexation by Cd, Zn, and Pb slightly decreased or did not significantly affect the oxidation rate of GSH, depending upon the pH (tested between pH 6 and 9). Thermodynamic modeling of GSH speciation demonstrated that the observed oxidation rates were not consistent with predicted free GSH3- concentration. These results indicated that Cd-, Zn-, and Pb-GSH complexes were susceptible to oxidation by a mechanism that differs from GSH3- oxidation. In contrast, Hg- and Ag-GSH complexes were inert for days, suggesting that they are stable for relatively long periods in the oxic water column. These results demonstrate that coordination of Hg(II) and Ag(I) to thiol-containing ligands can potentially increase persistence and transport in surface waters.
Introduction Low-molecular-weight thiols such as reduced glutathione (GSH), cysteine, and phytochelatins are produced by aquatic organisms during oxidative stress or exposure to toxic metals (1-4). GSH, in particular, is a ubiquitous polypeptide that is utilized by microorganisms and macrophytes as an antioxidant. Production of GSH can also reduce the toxicity of pollutant metals such as Hg, Ag, and Cd through complexation and reduction of bioavailable metal ions. Low-molecular-weight thiols are released by these organisms to the aquatic environment, resulting in concentrations ranging from 0.1 to 1000 nM in surface waters and sediment porewaters of aquatic systems (5-9). Thiols associated with natural organic matter (NOM) are also prevalent as a result of diagenesis processes, particularly in anaerobic systems (10). Humic substances from soil and aquatic environments contain reduced sulfur functional groups (oxidation state e0) at concentrations ranging between 20 and 140 µmol per g (11, 12). At these levels, thiols associated with NOM are likely to be controlling the aquatic speciation of trace metals such as Hg(II) and Ag(I). In surface waters, Hg(II) concentrations are generally less than 0.05 * Corresponding author phone: (919) 660-5109; fax: (919) 6605219; e-mail:
[email protected]. 2338
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 7, 2007
nM (13), much less than the concentration of thiols. Therefore, dissolved Hg(II) is likely to be coordinated to thiolcontaining humic substances (14-16). Free thiols (R-SH and R-S-) are unstable in the oxic aquatic environment. They can oxidize within minutes by reactive oxygen species such as H2O2 to form disulfide (RS-S-R’) and other organic sulfur oxidation products (1720). Dissolved Cu(II) and Fe(III)-oxide mineral surfaces can also catalyze the O2-oxidation of thiols (21-24). Therefore, free thiol ligands are not expected to persist in oxic environments. The oxidation rates of thiols are highest at pH values near or above the pKa of the thiol functional group (generally greater than pH 8) (20, 25). Thus, the oxidation reaction likely occurs through the deprotonated thiol (i.e., R-S-). Metals such as Hg(II), Ag(I), Cd(II), Zn(II), and Pb(II) can form stable complexes with thiols and potentially prevent oxidation by coordination to the thiolate. Previous studies have demonstrated that inorganic sulfide is stable in oxic waters when coordinated to metals such as Zn and Hg (26, 27). However, the extent of thiol stabilization by metals has not been closely considered. This knowledge could explain the persistence of thiols in surface waters and other environments where free thiols would otherwise oxidize. The stabilization of thiols by metal complexation can also interfere with the intracellular toxicity response systems of organisms. GSH is an important anti-oxidant that protects cells from oxidative stress induced by an excess of reactive oxygen species (21). Bacterial response mechanisms, such as those which regulate cell osmolarity and cytoplasmic pH, can be triggered by elevated levels of oxidized glutathione (GSSG) and other glutathione conjugates (28). This intracellular toxicity response system would be suppressed if complexation of GSH by metals reduced the formation of GSSG and other GSH conjugates. The goal of this study was to assess the stabilization of GSH in oxidizing conditions due to complexation by Hg, Ag, Cd, Zn, and Pb. Batch aqueous solutions of free GSH and metal-complexed GSH were exposed to either an excess of H2O2 or to small quantities of Cu2+, which induced catalytic oxidation of GSH. The studies indicated that complexation of GSH by trace metals resulted in decreased rates of oxidation. Hg- and Ag-GSH complexes were particularly inert during oxidation, suggesting that they are extremely stable in the aquatic environment.
Experimental Methods Materials. All chemicals and materials were purchased from Fisher Scientific, unless otherwise noted. Stock solutions were prepared with 18 MΩ-cm ultrapure water (Barnstead Nanopure). Acid-cleaned plasticware and glassware were used to contain and transfer all samples. Trace-metal-grade acids were used for pH adjustments. Stock solutions of 5.0 mM GSH (Sigma-Aldrich) were prepared and stored at 4 °C for less than 2 weeks. Metal stock solutions were prepared by dissolving 2.0-5.0 mM of Hg(NO3)2‚2H2O, AgNO3, Cd(NO3)2‚ 4H2O, Zn(NO3)2‚6H2O, Pb(NO3)2, and CuSO4 (Sigma Aldrich) in 0.1 M HNO3. Working solutions of 100 mM H2O2 were utilized within 1 h after dilution from a 30% H2O2 stock stored at 4 °C. Reagents for GSH derivatization included 2.0 mM 2,2′-dithiobis(5-nitropyridine) (DTNP) dissolved in acetonitrile and 0.5 M sodium acetate buffer adjusted to pH 6. Sealed ampules of 200 mM tributylphosphine (TBP) (Sigma Aldrich) were stored at 4 °C and utilized within 1 week after opening. Ammonium 7-fluorobenzo-2-oxa-1,3-diazole-4sulfonate (SBD-F) was dissolved at a concentration of 1 g per 4 mL and stored at 4 °C. 10.1021/es062269+ CCC: $37.00
2007 American Chemical Society Published on Web 02/21/2007
GSH Batch Solutions. Model freshwater solutions were prepared by diluting the GSH stock to 5.0 µM in an airsaturated buffered solution containing 0.01 M KCl. In the Cu(II)-oxidation studies, the pH was controlled by either 4 mM piperazine-1,4-bis(2-ethanesulfonic acid) (PIPES) or 4-(2-hydroxyethyl)piperazine-1-ethanesulfonic acid (HEPES) buffer (Sigma Aldrich). In the H2O2 oxidation experiments, 4.0 mM K2HPO4‚3H2O was the pH buffer. The pH was adjusted to between 6 and 9 by HCl or NaOH. In the metal-GSH models Cd, Zn, or Pb was added to a final concentration of 5 µM prior to equimolar addition of GSH. In the Hg and Ag models, the concentration was varied between 0.83 and 2.5 µM for Hg and between 1.25 and 5.0 µM for Ag. The model solutions were prepared in 50-mL screw cap polypropylene centrifuge tubes and were stored in the dark between sample collection times. Oxidation Studies. Oxidation of the GSH and metalGSH solutions were conducted in the presence of either H2O2 or Cu(II). In the H2O2 experiments, the oxidation reaction was initiated by adding H2O2 (diluted to reach 100 µM in the experimental model solutions). During the Cu(II)-catalyzed oxidation, the reaction was initiated by amending the model solutions with 0.2 µM Cu(II) final concentration. The concentration of GSH was monitored by taking 1-mL aliquots over time and stopping the oxidation reaction by adding 100 µL of the DTNP reagent (to achieve final concentration of 200 µM DTNP) and 50 µL of the 0.5 M sodium acetate stock. DTNP reacts with thiols to form a stable disulfide derivative (5). To identify the products of GSH oxidation, the concentration of GSSG was monitored in the GSH and metal-GSH models. During the course of the oxidation experiment, 250µL aliquots were diluted to a total of 500 µL with ultrapure water. TBP reducing agent (15 µL of 200 mM TBP) was then added to the sample and allowed to react for 10 min. The reduced GSH was then derivatized with SBD-F (40 µL of the stock), according to previously described methods (29). GSH Analysis. The samples containing derivatized GSH were analyzed by reverse phase high performance liquid chromatography (HPLC) within 36 h of derivatization. Standard calibrations were prepared daily and analyzed with the experimental samples. The DTNP-derivatives and SBDF-derivatives were quantified using previously described gradient programs (5, 29) on an HPLC system (Varian, Alltech Altima C18 column) and either UV absorbance (for DTNP method) or fluorescence detection (SBD-F method). The detection limit by the DTNP method was less than 0.15 µM GSH. The detection limit of the SBD-F method was less than 0.05 µM total glutathione. In a subset of the DTNP-derivatized samples, duplicate measurements were taken by re-injecting the derivatized samples 12-24 h after the first measurement. The second measurement averaged 89.2% of the first ((6.3%, n ) 10). Therefore, the DTNP-derivatization step was considered to sufficiently stop the oxidation process for HPLC analysis. To confirm the recovery of metal-complexed GSH by the DTNP and SBD-F reagents, aliquots of the metal-GSH samples were taken prior to addition of H2O2 or Cu(II). The average recovery of GSH (in the presence of Hg, Ag, Cd, Zn, and Pb) by DTNP-derivatization was 93.7% ((6.3%, n ) 48). In the samples that were reduced by TBP and subsequently derivatized by SBD-F, the resulting measurement of total glutathione includes both GSH and GSSG species. The average recovery of metal-complexed GSH by the SBD-F method prior to oxidation was 94.5% ((13.2%, n ) 17).
Results and Discussion Metal-GSH Oxidation Studies. The oxidation studies demonstrated that complexation by Hg and Ag prevented the complete oxidation of GSH during the 2-3 day time course
FIGURE 1. Oxidation of 5.0 µM GSH in the presence of 5 µM Cd ([), Zn (9), Pb (2), Ag (O), or 2.5 µM Hg (4) or without metals (×). (a) Oxidation by 100 µM H2O2, pH 7.8-7.9; (b) Oxidation by 0.2 µM Cu(II), pH 7.7-7.8. of the studies. Cd, Zn, and Pb slowed but did not prevent oxidation. For example at pH 7.8-7.9, the oxidation of metalfree GSH by excess H2O2 occurred within 2 h (Figure 1a). The observed first-order rate constant kobs, which was calculated utilizing data points collected in the first hour of the oxidation reaction, was 1.02 ((0.08) h-1. In the presence of equimolar concentrations of Cd, Zn, and Pb, the removal of GSH occurred at a slightly slower rate. Observed initial oxidation rates were 0.56 ((0.04) h-1 for Cd-GSH, 0.71 ((0.02) h-1 for Zn-GSH, and 0.73 ((0.03) h-1 for Pb-GSH. When the GSH solutions were amended with 2.5 µM Hg(II) or 5.0 µM Ag(I) to form Hg(GSH)2 and Ag(GSH) complexes, GSH was completely protected from oxidation for at least several hours (Figure 1a). At 4 days of exposure to H2O2, 85.8% of Hg(GSH)2 and 62.0% of Ag(GSH) was recovered (data shown in Supporting Information Figure S1), corresponding to observed rates that were less than 0.02 h-1. Therefore, complexation by Hg and Ag significantly slowed H2O2-oxidation of GSH. Similarly in the Cu(II)-catalyzed oxidation studies, Cd, Zn, and Pb slowed, but did not prevent the oxidation of GSH, whereas Hg and Ag protected GSH for days. For example at pH 7.7-7.8, trace quantities of Cu(II) resulted in oxidation of free GSH so that greater than 90% was removed within 1 h (Figure 1b). The observed oxidation rate, kobs, was 4.28 ((0.13) hr-1. The presence of Cd, Zn, and Pb slightly decreased the Cu(II)-catalyzed oxidation rate of GSH (Figure 1b). The observed oxidation rates were estimated to be 1.35 ((0.30) h-1 for Cd-GSH, 1.79 ((0.30) h-1 for Zn-GSH, and 3.80 ((0.13) h-1 for Pb-GSH. In the presence of Hg and Ag, GSH was stable for days. After 4 days of exposure to Cu(II), 98-104% of Hg(GSH)2 and Ag(GSH) complexes were recovered, corresponding to oxidation rates less than 0.001 h-1. VOL. 41, NO. 7, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2339
FIGURE 2. Observed oxidation rate of 5.0 µM GSH as a function of pH. Oxidation was initiated by (a) 100 µM H2O2; or (b) 0.2 µM Cu(II). Prior to oxidation, the GSH solutions contained the following metals: 5.0 µM Cd ([), Zn (9), Pb (2), Ag (O), or 2.5 µM Hg (4) or no metals (×). Error bars represent ( 1 standard error (n ) 4-8) of the estimated kobs. The observed rate of oxidation for metal-free GSH and metal-complexed GSH was dependent upon the pH of the model solutions (Figure 2a and 2b). The oxidation rate of free GSH and Pb-GSH complexes increased from pH 6-9 in both the H2O2 and Cu(II) oxidation studies. In the presence of Cd and Zn, the observed oxidation rates also increased with pH; however, above pH 7, estimated kobs values were less than those measured in the metal-free GSH solutions. In the presence of Hg and Ag, GSH was stable for days, regardless of pH. The experiments with Hg-GSH and Ag-GSH complexes indicated that these metals protected GSH from oxidation. To determine the stoichiometry of the stable GSH complexes, additional Cu(II)-catalyzed oxidation experiments were conducted at varying concentration ratios of Hg:GSH and Ag:GSH (Figure 3a and b). After approximately 24 h exposure to Cu(II), the stoichiometric ratio of Hg:GSH was 1.7 ((0.13) in all three mixtures. These results indicated that 1 mole of Hg protected 2 mol of GSH (Figure 3a). Because thiol concentrations are usually orders of magnitude greater than Hg(II) concentration in surface waters (11-13), these results indicated that Hg(II) can be extremely stable when coordinated to two thiols. In the Ag-GSH solutions, the Ag:GSH ratio was 0.83 ((0.13, n ) 3) after 24 h oxidation (Figure 3b), indicating that Ag persists as a 1:1 Ag:thiol complex. GSH Oxidation Products. Previous studies have suggested that oxidized glutathione (GSSG) is the major product of GSH oxidation (18, 21). In the oxidation experiments for free GSH, Cd-GSH, Zn-GSH, and Pb-GSH complexes, the concentration of the total glutathione (which includes both GSH and GSSG) was measured by TBP-reduction and SBD-F derivatization. In the H2O2 oxidation studies, the average total glutathione was 88.4% ((9.5%) of the initial GSH 2340
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 7, 2007
FIGURE 3. Oxidation of 5.0 µM GSH by 0.2 µM Cu in the presence of varying concentrations of (a) Hg and (b) Ag. (10 mM KCl, pH 7.3) concentration after 1.3 h of oxidation (Supporting Information Figure S2a). At this time period, GSH accounted for less than half of the total glutathione (Figure 1a), indicating that GSSG was the predominant oxidation product. In the Cuoxidation studies, GSSG was also a product of oxidation, as indicated by the difference between the total glutathione concentration (Supporting Information Figure S2b) and the GSH concentration in the uncomplexed GSH and Pb-GSH solutions (Figure 1b). However, GSSG did not appear to be stable in the presence of Cu(II). Only 42 and 71% of the initial GSH in the free GSH and Pb-GSH solutions, respectively, were recovered as total glutathione after 1 h. During the SBD-F derivatization method, the recovered glutathione in the Pb-GSH solutions was incomplete. Prior to the start of oxidation, only 80.6% ((17.2%, n ) 5) of total glutathione was recovered by the SBD-F method, whereas 97.3% ((3.2%, n ) 5) GSH was recovered in the same PbGSH solutions by the DTNP derivatization method. The discrepancy indicates that Pb may be interfering with the TBP reduction and SBD-F derivatization reactions. In contrast, the other metals did not appear to interfere with the derivatization methods. In the Cd-, Zn-, Hg-, and AgGSH solutions, the recovered glutathione prior to oxidation was 100.0% ((5.8%, n ) 7) by the SBD-F method and 93.7% ((4.0%, n ) 9) by the DTNP method. Mechanisms for Metal-GSH Oxidation. Previous studies have indicated that oxidation of GSH by H2O2 occurs through the thiolate functional group (i.e., GSH3-) (18, 20). The GSH oxidation reaction has been modeled in (20) using the following pathway and rate equation:
GSH3- + H2O2 f products
(1)
d[GSH] ) -k1[H2O2][GSH3-] dt
(2)
The rate constant k1 corresponds to the second-order reaction
the thiolate and Hg2+ likely resulted in a particularly inert complex that was not susceptible to H2O2 oxidation. Slow re-equilibration of GSH species in the presence of Hg(II) during the oxidation reactions would also reduce observed oxidation rates. The catalytic oxidation of GSH by Cu(II) occurs through a different pathway (21):
GSH3- + Cu2+ f CuGSH- f GSH•2- + Cu +
(4)
2GSH•2- f GSSG4-
(5)
Cu+ + O2 f Cu2+ + O•2
(6)
During Cu(II)-oxidation of the uncomplexed GSH, kobs increased as the GSH3- concentration increased (Figure 4b), indicating that the rate of oxidation was limited by the concentration of GSH3- for the steps described in eq 4. In the metal-GSH solutions an initial Cu2+exchange step, such as described in eq 7, was likely to occur prior to electron transfer between GSH and Cu(II):
MeGSHn-3 + Cu2+ f CuGSH- + Men+
FIGURE 4. Observed oxidation rate as a function of calculated GSH3- concentration prior to oxidation. Solutions contained 5 µM GSH and the following metals: 5 µM Cd ([), Zn (9), Pb (2), 2.5 µM Hg (4), or no metals (×). (a) Oxidation by 100 µM H2O2; (b) Oxidation by 0.2 µM Cu(II). Error bars represent ( 1 standard error of the estimated kobs. described by eq 1. In these experiments, H2O2 was in excess so that the calculated observed first-order rates kobs can be approximated by using the initial H2O2 concentration ([H2O2]o) and the following equation:
kobs ) k1[H2O2]o
[GSH3-] [GSH]T
(3)
If the overall oxidation rate was controlled by the availability of GSH3-, then the initial rate kobs should be consistent with the initial GSH3- concentration in both the metal-GSH and metal-free GSH solutions. The concentration of GSH3- in the metal-GSH solutions was calculated on MINEQL+ (30) using published GSH acidity constants and metal-GSH stability constants (31, 32) (summarized in Supporting Information Table S1). These calculations demonstrated inconsistencies in oxidation rates between the metal-GSH complexes and the metal-free GSH (Figure 4). In the Cd-, Zn-, and Pb-GSH solutions, kobs was greater than the corresponding kobs values in the metal-free GSH solution (Figure 4a). In contrast, kobs for Hg(GSH)2 complexes was significantly lower. The dependence of kobs on pH (as shown in Figure 2) was not only a function of GSH3- concentration in the metal-GSH solutions, but also on the relative stability of the metal-GSH complex. Cd-GSH, Zn-GSH, and Pb-GSH complexes may be susceptible to oxidation if coordination to GSH occurs through multiple functional groups such as the amine and carboxylate groups. Chelation of these metals could expose the thiolate to nucleophilic attack by H2O2. Hg(II) generally has a much higher affinity for thiolate over carboxylate and amine functional groups. Therefore, coordination between
(7)
In the Zn-GSH solutions, kobs values were slightly less than or approximately equal to kobs in the metal-free GSH oxidation (Figure 4b). In the Pb-GSH solutions, kobs was equal to or greater than kobs in the metal-free control (with exception to the first data point in Figure 4b). These results indicated that the stabilities of Zn-GSH and Pb-GSH complexes were not limiting factors during the Cu2+ exchange step described by eq 7. In the Hg-GSH solutions, kobs (
