Chem. Res. Toxicol. 1996, 9, 709-712
709
The Formation Constants of Mercury(II)-Glutathione Complexes Paul D. Oram, Xiaojun Fang,† and Quintus Fernando* Department of Chemistry, University of Arizona, Tucson, Arizona 85721
Peter Letkeman and Douglas Letkeman Department of Chemistry, Brandon University, Brandon, Manitoba, Canada R7A 6A9 Received November 13, 1995X
The formation constants of the 1:1 and 1:2 complexes of Hg(II) with glutathione and their protonated species have been determined by using a competitive potentiometric titration with the competing ligand diethylenetriaminepentaacetic acid (DTPA). The formation constants of the 1:1 complex and its protonated species have not been reported previously. The formation constant of the 1:2 complex of Hg(II) and glutathione is substantially smaller than the accepted value that has been reported in the literature. These results have important implications in the models that have been employed to explain the mobilization and distribution of Hg(II) in biological systems.
Introduction Numerous studies have shown that Hg(II) binds strongly to ligands with free mercapto groups. In biological systems, the activity of intra- and extracellular proteins and enzymes with free mercapto groups is adversely affected by binding of the -SH groups to Hg(II). The tripeptide glutathione, which is found in approximately millimolar concentrations in the erythrocytes of whole blood, has several important functions. One such function is to maintain the -SH groups in certain proteins, which are essential for their activity, in a reduced state by preventing the oxidation of the -SH groups to disulfide groups. An understanding of the nature and the extent of binding of glutathione to Hg(II) is, therefore, of importance in seeking explanations for the manifestation of the acute and chronic toxic effects of Hg(II). Glutathione has been used as an antidote for mercury poisoning, but without much success. A knowledge of the magnitude of the formation constants of the Hg(II)-glutathione complexes is essential for the development of effective antidotes for mercury poisoning. These formation constants are also important in understanding the manner in which Hg(II) is mobilized and transported in biological systems. There is a serious problem, however, with the published value of the formation constant of the 1:2 mercury(II)-glutathione complex, HgL2, where L represents the completely deprotonated glutathione molecule. All charges are omitted from the ligand and the mercury complexes for the sake of simplicity. There is also some inconsistency in the published reports on the stoichiometry of the Hg(II)glutathione complexes that are formed at various pH values and at various Hg(II):glutathione ratios. In a pioneering polarographic investigation of the reaction between Hg(II) and glutathione, Stricks and Kolthoff (1) showed that the complexes HgL2, Hg2L2, and Hg3L2 were formed in the pH range 3-9 in the absence of chloride * To whom correspondence should be addressed. † Department of Pharmacology and Toxicology. X Abstract published in Advance ACS Abstracts, April 15, 1996.
S0893-228x(95)00189-5 CCC: $12.00
ions. In the presence of a high concentration of chloride ions, only the species HgL2 and HgCl42- were formed; there was no evidence for the formation of the complex HgL. The formation constants of the species HgL2 and the protonated species HgHL2 and HgH2L2 were obtained by calculating the free Hg(II) concentration from the measured value of the potential of a mercury electrode vs a reference saturated calomel electrode. Potentiometric titrations of mixtures of Hg(II) and glutathione in varying ratios were performed with a standard solution of NaOH by Kapoor, Doughty, and Gorin (2). They concluded from the shapes of the titration curves that the same species, HgL2, Hg2L2, and Hg3L2, reported by Stricks and Kolthoff (1) were formed in solution, but no evidence was found for the presence of the species HgL in solution. In the complex HgL2, only the sulfur atoms in the mercapto groups are coordinated to Hg(II). This was convincingly demonstrated by Fuhr and Rabenstein (3), who monitored the chemical shifts of the carbon atoms in the coordinated glutathione molecules by 13C NMR in solutions containing varying ratios of Hg(II): glutathione. A 1:1 complex, HgL, was isolated in the form of a precipitate from an aqueous ethanol solution by Neville and Drakenburg (4), and its structure in solution was deduced by 13C NMR to be a chelate in which the donor atoms are the mercapto sulfur atom and the nitrogen atom in the glycine residue of glutathione. The presence of the complex HgL in acidic solution was confirmed by Katojno, Inoue, and Chuˆjoˆ (5), who also employed 13C NMR. Their results, however, indicated that the donor atoms in the chelate ring were the mercapto sulfur atom and the carbonyl oxygen atom in the cysteinyl residue of glutathione. In a recent 13C NMR study by Cheesman, Arnold, and Rabenstein (6), evidence for the formation of an additional complex, HgL3, was obtained in solutions containing a glutathione:Hg(II) ratio in excess of 2:1 at physiological pH. On the basis of this work, it was proposed that, although the thermodynamic stabilities of the Hg(II)-glutathione complexes are very high, the bonds formed between Hg(II) and -SH groups are labile, and the Hg(II) is continually exchang© 1996 American Chemical Society
710 Chem. Res. Toxicol., Vol. 9, No. 4, 1996
ing -SH groups in the glutathione ligands. Shoukry, Cheesman, and Rabenstein (7) used a completely independent method, a polarimetric method, to confirm the results obtained in the 13C NMR study. The magnitude of the formation constant of the 1:3 complex, HgL3, formed from HgL2 and the ligand, L, indicated that the third ligand was much more loosely bound than the first two ligands. In addition, the stepwise displacement constants of the ligand, L, in HgL2 by penicillamine and 2-mercaptoethylamine were determined, and it was shown that the relative stabilities of the 1:2 complexes of Hg(II) with these two ligands containing mercapto groups was dependent on the magnitude of their protonation constants. All the previous work on the stoichiometry and the formation constants of Hg(II)-glutathione complexes can be summarized as follows: the complexes HgL, Hg2L2, HgL2, Hg3L2, and HgL3 as well as several protonated complexes are formed in aqueous solutions at varying pH and varying Hg(II):glutathione ratios. To date, no attempt has been made to confirm the extraordinarily high formation constant of HgL2 that was determined with a mercury electrode (1). Casas and Jones (8) described the difficulties that were encountered in their attempts to rationalize the formation constants of Hg(II) complexes formed with ligands bearing one or more mercapto groups. In our work with ligands that have two vicinal mercapto groups, meso-dimercaptosuccinic acid (meso-DMSA)1 and racemic-dimercaptosuccinic acid (rac-DMSA) (9), we have measured the formation constants of their mercury(II) complexes by employing a competitive potentiometric titration. This method was first used by Ackermann and Schwarzenbach (10) for the determination of the formation constants of transition metals with aminocarboxylic acids. They recognized that it is very difficult to determine formation constants for strong complexes by direct titration, because the equilibrium is shifted so far to the right that it becomes virtually impossible to measure accurately the concentrations of the reactants. In order to solve this problem, they performed potentiometric titrations in the presence of ligands that could compete with the aminocarboxylic acid under investigation. The success of the method is evidenced by the confirmation of their values of metalEDTA formation constants in many subsequent studies. In our work that is reported below, we have measured the formation constants of Hg(II)-glutathione complexes by employing a similar competitive potentiometric titration with diethylenetriaminepentaacetic acid (DTPA) as a competing ligand.
Experimental Section Caution: Mercury compounds and concentrated acids should be handled with gloves in a fume hood. Materials. The glutathione (98-100%) was purchased from Sigma Chemical Co. (St. Louis, MO) and was used without further purification. The ligand diethylenetriaminepentaacetic acid (DTPA) was purchased from Aldrich Chemical Co. (Milwaukee, WI) or J. T. Baker (99.5%) (Phillipsburg, NJ) and used without further purification. Disodium ethylenediaminetetraacetic acid (EDTA) (Certified ACS) was purchased from Fisher (Pittsburgh, PA). All metal and salt solutions were prepared from ACS reagent grade materials. Titrations were carried out at Brandon University with potassium hydroxide prepared from “Dilut-It” concentrate, purchased from J. T. Baker (Phillipsburg, 1 Abbreviations: meso-DMSA, meso-dimercaptosuccinic acid; racDMSA, racemic-dimercaptosuccinic acid; DTPA, diethylenetriaminepentaacetic acid.
Oram et al. NJ). Titrations were performed at the University of Arizona with potassium hydroxide purchased from Mallinckrodt (Chesterfield, MO). Concentrated hydrochloric and nitric acids, Ultrex, were purchased from J. T. Baker (Phillipsburg, NJ), and standard 0.1000 M hydrochloric acid was purchased from Ricca Chemical Co. (Arlington, TX). Potassium hydrogen iodate (99.9+%) was purchased from Aldrich Chemical Co. (Milwaukee, WI). Ultrapure DNA grade agarose was purchased from BioRad Laboratories (Hercules, CA), and the standard pH 4.00 and 10.00 buffer solutions were from Fisher (Pittsburgh, PA). Preparation and Standardization of the Solutions. EDTA solutions (0.05 M) were standardized with a standard zinc solution, with xylenol orange as indicator, and the concentration of a 0.01 M stock solution of DTPA was determined by titration with a standard lead solution, with xylenol orange as indicator (11). The DTPA solution was made in 0.1 M KNO3 in order maintain a constant ionic strength in the titration solutions. Stock solutions of mercuric nitrate and mercuric chloride (0.05 M) were prepared in distilled, deionized water. It was necessary to add a small amount of concentrated nitric acid to completely dissolve the mercuric nitrate. These solutions were standardized either directly with EDTA, or by backtitration of excess EDTA with a standard zinc solution, with xylenol orange as indicator (11). The excess acid in the mercuric nitrate solution was determined by adding a known amount of EDTA that was equimolar or slightly in excess of the mercury concentration, and then performing a potentiometric titration with a solution of standard base (12). Glutathione stock solutions that were approximately 0.05 M were made daily in freshly boiled, distilled deionized water, from which oxygen was removed by bubbling nitrogen through the solution. Approximately 0.1 mmol of the ligand was titrated with standard base, and the program PKAS (13, 14) was used to determine the concentration of the ligand, as well as the acid dissociation constants of the ligand. Standard HCl (0.1000 M) was added to the titration vessel to adjust the initial pH to approximately 2.8. A 50% w/v stock solution of potassium hydroxide was made in freshly boiled distilled deionized water and filtered through a GF/A glass fiber filter purchased from Gelman (Ann Arbor, MI). This solution was used to prepare a 0.1 M KOH solution in freshly boiled, distilled deionized water which had been cooled under nitrogen. The base solution was kept under nitrogen in the reservoir of a 5 mL buret and was standardized against potassium hydrogen iodate with methyl red as indicator. Solutions of 0.1 M KCl and 0.1 M KNO3 were made in freshly boiled, distilled deionized water. Potentiometric Determinations. Potentiometric titrations at the University of Arizona were carried out with the aid of a Beckman (Fullerton, CA) Model Phi 72 pH meter with a Corning (Corning, NY) 476280 glass electrode and Corning 476350 calomel reference electrode. A 1 mm inner diameter Teflon tube filled with agarose, saturated with KNO3, was used as a salt bridge between the reference electrode and the titration vessel. Titrations at Brandon University were carried out with the aid of a Fisher (Pittsburgh, PA) Accumet #50 pH meter, a Sargent (Skokie, IL) glass-calomel combination electrode, and a Metrohm (Herisau, Switzerland) 10 mL microburet. Titrations were performed in a sealed, jacketed vessel under nitrogen at 25 ( 0.1 °C. All titrations were carried out at an ionic strength of 0.1, with either potassium chloride or potassium nitrate as a background electrolyte. The slope of the pH meter response was adjusted using standard pH 4.00 and 10.00 buffers. Calibration of the pH meter was performed by titrating a 0.1000 M standard hydrochloric acid solution with the standard KOH solution. Linear plots of the measured pH versus calculated -log[H] in the acidic and basic regions were used to convert measured pH values to hydrogen ion concentrations. A Gran plot was used to estimate the amount of carbonate contamination in the base, which was usually between 0.5% and 1% (15-17). A Gran plot of the calibration data in the basic region was also used to determine an experimental value for the ion product for water at the ionic strength employed in the potentiometric titrations. Metal-ligand formation constants were determined as follows. Solutions containing only the mercuric ion and glu-
Mercury(II)-Glutathione Formation Constants
Chem. Res. Toxicol., Vol. 9, No. 4, 1996 711
Table 1. Experimentally Determined Glutathione Acid Dissociation Constants and Literature Values (25 °C, Ionic Strength ) 0.1) pK1 avgsa avgsb
2.08 ( 0.10 2.16 ( 0.10
ref 18 ref 19 ref 20 avgs
2.12 2.09 1.98 2.06
pK2
pK3
Experimental Values 3.50 ( 0.02 8.68 ( 0.03 3.55 ( 0.05 8.65 ( 0.05 Literature Values 3.53 8.66 3.48 8.67 3.49 8.75 3.50 8.69
pK4 9.54 ( 0.02 9.56 ( 0.05 9.62 9.54 9.69 9.62
a The uncertainties are for the average of seven experiments at a 95% confidence level (University of Arizona). b The uncertainties are for the average of five experiments at a 95% confidence level (Brandon University).
tathione in various ratios were titrated potentiometrically with a standard solution of KOH in the presence of 0.1 M KCl as a background electrolyte. The known formation constants of the chloro complexes of Hg(II) were included in all calculations. Under most conditions the chloride ion does not compete directly with glutathione, but it does serve an important function in keeping any free mercury(II) from being reduced. The competitive potentiometric titrations were carried out in the presence of 0.1 M potassium nitrate as the background electrolyte. Although a white precipitate was observed initially upon addition of the mercury(II) to the solution containing the glutathione and the competing ligand DTPA, it dissolved immediately upon mixing. A series of experiments were carried out in which the ratios of DTPA, glutathione, and Hg(II) were varied. In one experiment, equimolar amounts of Hg(II), glutathione, and Na2EDTA were used. With few exceptions, stable pH readings were attained in less than 1 min. At selected points during the titration, 5-10 min was allowed for equilibrium to be attained, and the pH readings remained constant during these periods. When competing ligands were used, only one or two data points near neutrality required longer equilibration periods, and this was often greater than 1 h. In these cases the pH meter reading was taken when the meter drift was less than 0.002 pH unit per 5 min. The formation constants of the Hg(II) complexes were calculated by the program BEST (14), which uses the Newton-Raphson method of solving simultaneous equations, and a nonlinear least-squares curve fitting routine.
Results and Discussion The average pKa values of glutathione calculated from nine potentiometric titrations performed at the University of Arizona and from five titrations performed at Brandon University are collected in Table 1, together with pKa values reported in the literature for comparison. There is excellent agreement between these independently determined values. The formation constants of the Hg(II)-glutathione complexes were calculated from a series of potentiometric titration curves, on the assumption that the initial concentrations of Hg(II), glutathione, and the competing ligand could be accounted for by the various species that were formed in solution. In a series of initial experiments in the absence of competing ligand, the solution composition was modeled by assuming that only the 1:2 species, HgL2, was present and that the overall formation constant was 1042 (1). This model gave a very poor fit of the experimental data, even after HgL and the protonated complexes, HgLH, HgLH2, HgL2H, and HgL2H2, were included in the model. In the presence of the competing ligand, DTPA, the best model of the solution composition included both the 1:1 and 1:2 complexes, HgL and HgL2, as well as their protonated species, but with much lower formation constants. This new model also resulted in a
Table 2. Logarithm of Mercury-Glutathione Formation Constants (25 °C, Ionic Strength ) 0.1)a species log βpqrb
p
q
r
1 1 1 1
1 1 1 1
0 1 2 -1
26.04 (0.95)c 32.49 (0.79)c 35.68 (1.03)c 15.80 (1.98)c
1 1 1 1
2 2 2 2
0 1 2 3
33.40 (1.03)c 42.40 (2.11)c 52.29 (0.85)c 55.28 (1.15)c
41.58d 41.92d 40.96d
a For experiments involving chloride ion as a background electrolyte, the log β values used in the calculations were as follows: HgCl+ ) 6.74, HgCl2(aq) ) 13.22, HgCl3- ) 14.1, and HgCl42- ) 15.1 (21). b βpqr ) [(MpLqHr)2p-3q+r]/[M2+]p[L3-]q[H+]r, L3- represents the completely deprotonated glutathione. c Number in parentheses is standard deviation for the average value of 20 experiments. The σ(pH)fit values calculated by the BEST program for the individual experiments ranged from 0.002 to 0.01. d Value reported by Stricks and Kolthoff (1).
Figure 1. Species distribution diagram for a solution containing 0.001 M glutathione and 0.0005 M mercury(II).
good fit for the data from titrations where no competing ligand was used. The average values of the formation constants of the Hg(II)-glutathione complexes calculated from 20 potentiometric titrations performed at the University of Arizona and at Brandon University are shown in Table 2. The σ(pH)fit values for these experiments ranged between 0.002 to 0.01. In the algebraic expression for the equilibrium constant, βpqr, a value of r ) -1 implies that a hydroxy species, e.g., ML(OH) is formed. The species distribution diagrams, calculated on the basis of the formation constants in Table 2, are shown in Figures 1, 2, and 3. In a solution containing a Hg(II):glutathione ratio of exactly 1:2, the principal species, in the pH range 4-8, is HgL2H2; above pH 8.5, the three species HgL2H2, HgL2H, and HgL2 are all present in significant concentrations. If, however, the Hg(II):glutathione ratio varies slightly from the value of 1:2, the species distribution diagram (Figure 2) shows several important differences from the species distribution diagram in Figure 1. In addition to the principal species, HgL2H2, the species HgLH and HgL are also present, each at about 10% of the total concentration of the glutathione. Previous workers did not report the presence of these 1:1 Hg(II):complexes, probably because they are present in relatively low concentrations. In solutions containing a 1:1 ratio of Hg(II):glutathione (Figure 3) the situation is quite different. The 1:1 complexes HgLH2, HgLH, and HgL are present in high concentrations
712 Chem. Res. Toxicol., Vol. 9, No. 4, 1996
Oram et al.
that the reported value of the formation constant of the 1:2 Hg(II)-glutathione complex (1) is incorrect.
Acknowledgment. Work done at the University of Arizona was supported by Grant ES04940 from the National Institute of Environmental Health Sciences, NIH. Work done at Brandon University was supported by a BURC grant. Supporting Information Available: Supporting information containing the summary of titration conditions (11 pages) is available. See any current masthead for ordering and Internet information.
References Figure 2. Species distribution diagram for a solution containing 0.000925 M glutathione and 0.000503 M mercury(II).
Figure 3. Species distribution diagram for a solution containing 0.001 M glutathione and 0.001 M mercury(II).
between pH 3.5 and 10.0, whereas the 1:2 complex HgL2H2 is present at a relatively low concentration (10% or less) in this pH range. It is evident, therefore, that the type of mercury(II)-glutathione complex that is present in solution depends on the pH of the solution, as well as the initial ratio of Hg(II):glutathione. The accuracy of the formation constants determined for species that have a low concentration in solution are inherently lower than the major species. In Table 2 there is a serious discrepancy between the previously reported value of the formation constant of the 1:2 species, HgL2, and the formation constant calculated from the competitive potentiometric titrations that we have carried out. In our work, pH was the experimental variable that was measured in the range 2 < pH < 11. A direct measurement of the free Hg2+ concentration in these solutions is not possible because the very high formation constants of the Hg(II)-glutathione complexes result in very low free Hg2+ concentrations that cannot be measured experimentally. The widely accepted value of 1041.58, of the formation constant of HgL2, was obtained by measuring the potential of a mercury electrode vs a reference saturated calomel electrode and employing the Nernst equation to calculate the free Hg2+ concentration, which was in the range of 10-42-10-44 M (1). The potential of the mercury electrode was not governed by these very low concentrations of free Hg2+, but was probably governed by the ions of the other electrolytes in solution that were present at much higher concentrations; moreover, it is unlikely that the measured potential of the mercury electrode was reversible. In view of these experimental difficulties, it is evident
(1) Stricks, W., and Kolthoff, I. M. (1953) Reactions between mercuric mercury and cysteine and glutathione. Apparent dissociation constants, heats and entropies of formation of various forms of mercuric mercaptocysteine and -glutathione. J. Am. Chem. Soc. 75, 5673-5681. (2) Kapoor, R. C., Doughty, G., and Gorin, G. (1965) The reaction and assay of glutathione with Hg2+ and alkali. Biochim. Biophys. Acta 100, 376-383. (3) Fuhr, B. J., and Rabentstein, D. L. (1973) Nuclear magnetic resonance studies of the solution chemistry of metal complexes. IX. The binding of cadmium, zinc, lead, and mercury by glutathione. J. Am. Chem. Soc. 95, 6944-6950. (4) Neville, G. A., and Drakenberg, T. (1974) Mercuric mercury and methylmercury complexes of glutathione. Acta Chem. Scand. B 28, 473-477. (5) Katojno, Y., Inoue, Y., and Chuˆjoˆ, R. (1977) Nuclear magnetic resonance studies of the complex formation of mercuric chloride with L-cysteine and with glutathione. Polymer J. 9, 471-478. (6) Cheesman, B. V., Arnold, A. P., and Rabenstein, D. L. (1988) Nuclear magnetic resonance studies of the solution chemistry of metal complexes. 25. Hg(thiol)3 complexes and Hg(II)-thiol ligand exchange kinetics. J. Am. Chem. Soc. 110, 6359-6364. (7) Shoukry, M. M., Cheesman, B. V., and Rabentstein, D. L. (1988) Polarimetric and nuclear magnetic resonance studies of the complexation of mercury by thiols. Can. J. Chem. 66, 3184-3189. (8) Casas, J. S., and Jones, M. M. (1980) Mercury(II) complexes with sulfhydryl containing chelating agents: stability constant inconsistencies and their resolution. J. Inorg. Nucl. Chem. 42, 99102. (9) Fang, X., Hua, F., and Fernando, Q. (1996) Comparison of racand meso-2,3-dimercaptosuccinic acid (DMSA) for mobilizing mercury and cadmium using chemical speciation models. Chem. Res. Toxicol. 9, 284-290. (10) Ackermann, H., and Schwarzenbach, G. (1949) Complexones XVI. The Determination of Formation Constants of Highly Stable Complexes of the Iminodiacetic Acid Derivatives. Helv. Chim. Acta 32, 1543-1554. (11) Prˇibil, R. (1982) Applied Complexometry, Pergamon Press, Oxford. (12) Harris, W. R., and Martell, A. E. (1976) Aqueous complexes of gallium(III). Inorg. Chem. 15, 713-720. (13) Motekaitis, R. J., and Martell, A. E. (1982) BESTsA new program for rigorous calculation of equilibrium parameters of complex multicomponent systems. Can. J. Chem. 60, 2403-2409. (14) Martell, A. E., and Motekaitis, R. J. (1988) Determination and Use of Stability Constants, VCH, New York. (15) Gran, G. (1950) Determination of the equivalent point in potentiometric titrations. Acta Chem. Scand. 4, 559-577. (16) Gran, G. (1952) Determination of the equivalence point in potentiometric titrations. Part II. Analyst 77, 661-671. (17) Rossotti, F. J. C., and Rossotti, H. (1965) Potentiometric titrations using Gran plots. J. Chem. Educ. 42, 375-378. (18) Rabenstein, D. L. (1973) Nuclear magnetic resonance studies of the acid-base chemistry of amino acids and peptides. I. Microscopic methylmercury-complexed glutathione. J. Am. Chem. Soc. 95, 2797-2803. (19) Pillai, L., Boss, R. D., and Greenberg, M. S. (1979) On the role of solvent in complexation equilibria. II. The acid-base chemistry of some sulfhydryl and ammonium containing amino acids in water-acetonitrile mixed solvents. J. Solution Chem. 8, 635646. (20) Arnold, A. P., and Canty, A. J. (1983) Methylmercury(II)sulfydryl interactions. Potentiometric determination of the formation constants for complexation of methylmercury(II) by sulfhydryl containing amino acids and related molecules, including glutathione. Can. J. Chem. 61, 1428-1434. (21) Kotrly´, S., and Sˇ u˚cha, L. (1985) Handbook of Chemical Equilibria in Analytical Chemistry, Ellis Horwood Ltd., Chichester.
TX9501896
