Cadmium Uptake by Caco-2 Cells: Effects of Cd Complexation by Chloride, Glutathione, and Phytochelatins

Toxicology and Applied Pharmacology 170, 29 –38 (2001) doi:10.1006/taap.2000.9075, available online at http://www.idealibrary.com on

Cadmium Uptake by Caco-2 Cells: Effects of Cd Complexation by Chloride, Glutathione, and Phytochelatins Catherine Jumarie,* ,1 Claude Fortin,† Mario Houde,* Peter G. C. Campbell,† and Francine Denizeau‡ *De´partement des Sciences Biologiques, Universite´ du Que´bec a` Montre´al, C.P. 8888, Succ. Centre-ville, Montre´al, Que´bec, Canada, H3C 3P8; †INRS-Eau, Universite´ du Que´bec, C.P. 7500, Sainte-Foy, Que´bec, Canada, G1V 4C7; and ‡De´partement de Chimie, Universite´ du Que´bec a` Montre´al, C.P. 8888, Succ. Centre-ville, Montre´al, Que´bec, Canada, H3C 3P8 Received June 16, 2000; accepted September 19, 2000

(Kojima et al., 1985; Moberg Wing, 1993; Berglund et al., 1994). It is generally thought that this reduced bioavailability is the result of complexation by organic ligands in the bolus. A convincing body of data has accrued, largely in the field of aquatic toxicology, to support the idea that the biological response elicited by a dissolved metal is generally a function of the free-metal ion concentration (aquo-ion), a concept usually referred to as the “free-ion activity model” (Morel, 1983). Since to be biologically available an ingested metal should be absorbed into the blood circulation, it is thus often believed that the extent of metal absorption is directly dependent on the free-metal ion levels in the intestinal lumen. However, there are a number of examples where metal bioavailability diverges from predictions of the free-ion activity model (Campbell, 1995). One of the ligands of particular interest for its protective effect on heavy metal toxicity is glutathione. Glutathione (␥Glu-Cys-Gly) is the lowest-molecular-weight peptide in mammalian cells that contains a free thiol group, and its protective role against cellular oxidative damage is well recognized. Moreover, GSH is largely absorbed in the upper jejunum; a ␥-glutamyltranspeptidase-dependent system of absorption has been well characterized, but transport of intact GSH has also been demonstrated (Vincenzini et al., 1989; Hagen et al., 1990; Iantomasi et al., 1997). Luminal GSH is believed to participate in the detoxification of reactive electrophiles in the diet (Hagen et al., 1990). Glutathione forms complexes with Cd in vitro (Perrin and Watt, 1971) and it has been shown that exogenous GSH may decrease Cd uptake (Kang, 1992). This extracellular protection was attributed to the binding of Cd by GSH, thus reducing the free Cd 2⫹ concentration; it was assumed that the Cd–GSH complex was not transported. Unfortunately, metal speciation in these studies was not well controlled; the impact of speciation changes may have been misevaluated and estimation of Cd–GSH uptake becomes problematic. In contrast to the situation with GSH, intestinal absorption of phytochelatins has not been directly investigated. Phytochelatins (PCs) are glutathione-derived peptides with the general structure (␥-Glu-Cys) n Gly, where n varies from

Cadmium Uptake by Caco-2 Cells: Effects of Cd Complexation by Chloride, Glutathione, and Phytochelatins. Jumarie, C., Fortin, C., Houde, M., Campbell, P. G. C., and Denizeau, F. (2001). Toxicol. Appl. Pharmacol. 170, 29 –38. Short-term cadmium uptake by the highly differentiated TC7 clone of enterocytic-like Caco-2 cells was studied as a function of Cd speciation. For low metal concentrations and with a constant free [Cd 2ⴙ] ⴝ 43 nM, initial uptake rates of 109Cd increased linearly as a function of increasing concentration of chlorocomplexes (⌺[ 109CdCl n2ⴚn]) over the range from 0 to 250 nM. When normalized as a function of the metal concentration, the absorption rate for the chlorocomplexes was less than that estimated for uptake of the free Cd 2ⴙ cation. Metal absorption decreased upon organic ligand addition in the exposure media, but much less than predicted from the assumption that only inorganic metal species would be transported. Under exposure conditions where the concentration of each of the inorganic species was kept constant, 109Cd uptake increased with increasing concentrations of cadmium glutathione ( 109Cd–GSH) or phytochelatin ( 109Cd– hmPC 3) complexes. A specific system of very high affinity but low capacity has been characterized for 109Cd–GSH transport, whereas accumulation data increased linearly with 109Cd– hmPC 3 up to 6 ␮M. Comparison among uptake data for 0.3 ␮M inorganic 109Cd, 109Cd—GSH, or 109Cd– hmPC 3 yields the following accumulation ratios: Cd– GSH/Cd inorg ⴝ 0.2; Cd– hmPC 3/Cd inorg ⴝ 0.5. These results clearly show that Cd 2ⴙ is not the exclusive metal species participating in Cd absorption, though, for comparable Cd concentrations, its contribution to transport would be more important than that of other species. Cadmium bound to thiol-containing peptides may be absorbed via transport systems that differ from those involved in absorption of the inorganic metal species. © 2001 Academic Press Key Words: metal speciation; ligands; kinetics; intestinal cells.

It is now well recognized that the speciation of a metal may affect its absorption through the intestinal tract in various species, including humans, and several studies have shown that complexation with organic ligands reduces Cd absorption 1

To whom correspondence should be addressed. Fax: (514) 987-4647; E-mail: [email protected]. 29

0041-008X/01 $35.00 Copyright © 2001 by Academic Press All rights of reproduction in any form reserved.

30

JUMARIE ET AL.

2 to 11. These compounds, considered to be class III metallothioneins (Kojima, 1991), were first characterized in the fission yeast Schizosaccharomyces pompe and termed cadystins (Murasugi et al., 1981). They were subsequently purified from plant cells and termed phytochelatins; they are the major metal-binding ligands in contaminated plants (Grill et al., 1985). In addition to GSH, two other terminal tripeptides have been identified: ␥-Glu-Cys-Glu in maize (Meuwly et al., 1993) and ␥-Glu-Cys-Ser, specifically expressed in some Poaceae species (Klapheck et al., 1992). These latter serine-containing tripeptides have been called hm-GSH (Bergmann and Rennenberg, 1993); therefore the peptides (␥-Glu-Cys) n -Ser are designated hm-PC n . Concentrations of hm-PC 3 have been shown to be very sensitive to Cd exposure; the synthesis of this plant peptide is highly stimulated by Cd exposure in both shoots and roots of the Poaceae (Klapheck et al., 1994). In studies on rats, ingested Cd bound to PCs was found to be much less efficiently absorbed than inorganic Cd, administered in the diet as CdCl 2 ; Cd distribution was also different, with a higher ratio of renal Cd to hepatic Cd being observed with the phytochelatins (Fujita et al., 1993). However, no studies were conducted under well-controlled metal speciation conditions and, to date, no data are available on the intestinal absorption of Cd–PCs per se. Improved knowledge of the intestinal fate of these metal complexes is of primary interest, since substantial levels of Cd have been detected in some crops as well as in food products derived from contaminated plants, e.g., bread (Tahvonen and Kumpulainen, 1994). The goal of this study was thus to evaluate the influence of Cd speciation on Cd uptake by intestinal cells by i) estimating the relative contribution of the free Cd 2⫹ ion to total metal uptake from an aqueous solution and ii) measuring specifically the uptake of Cd bound to GSH or PCs. Our previous studies have shown that the highly differentiated TC7 clone of the enterocytic-like Caco-2 cells is a relevant in vitro model for studying the intestinal absorption of Cd; an apparent in-to-out metal accumulation ratio of 12 was measured following a 1-h exposure to 0.3 ␮M 109 Cd and a specific transport system, obeying Michaelis–Menten kinetics with a K m of 3.8 ⫾ 0.7 ␮M, has been characterized (Jumarie et al., 1997). More recently we demonstrated that Cd transport through Caco-2 cell monolayers involves mainly a transcellular route, the paracellular pathway being negligible over a 12-h exposure to nontoxic metal concentrations (Jumarie et al., 1999). These results have been further supported by Blais et al. (1999). The first step of intestinal Cd absorption is transport into the enterocyte through the apical membrane. The influence of Cd speciation on this transmembrane process has thus been investigated using TC7 cells grown on petri dishes.

METHODS Cell Cultures The TC7 clones, isolated from late passages of the Caco-2 cell line (Chantret et al., 1994), were kindly supplied by Dr. A. Zweibaum et al. (INSERM U178, Villejuif, France). The cells were used between the 50th and 58th passages. Stock cultures were grown in 75-cm 2 plastic flasks in Dulbecco’s modified Eagle essential minimum medium containing 25 mM glucose and supplemented with 15% inactivated (56°C for 30 min) fetal bovine serum (FBS), 0.1 mM nonessential amino acids, and 50 units/ml penicillin ⫹ 50 ␮g/ml streptomycin. Cells were maintained in a 5% CO 2–95% air atmosphere and the medium was changed every 2 days. Cultures were passaged weekly by trypsinization (0.05% trypsin– 0.53 mM EDTA). For all experiments, cells were seeded in 35 ⫻ 10 mm petri dishes at a density of 12 ⫻ 10 3 cells/cm 2. Cell confluence was achieved on day 7 and cell cultures were maintained for 3 weeks to reach the stationary growth phase and to allow functional differentiation (Chantret et al., 1994; Jumarie et al., 1999). Cadmium Accumulation and Transport Measurements Standard experimental conditions. Cadmium accumulation measurements were performed on 21-day-old cell cultures in serum-free transport medium, referred to as the standard transport medium in our previous studies (Jumarie et al., 1997, 1999). This medium contained the following (in mM): 137 NaCl, 4.7 KCl, 1.2 MgSO 4, 2.5 CaCl 2, 4 D-glucose, and 10 Hepes, buffered to pH 7.3 with 5 mM NaOH. Transport media were always prepared in advance and allowed to equilibrate overnight at room temperature. Cell monolayers were washed four times with the Cd-free transport medium, to eliminate culture medium and especially serum, prior to incubation in 1 ml transport medium containing 109Cd-labeled CdCl 2 (sp act 0.425 to 0.575 mCi/␮mol) at room temperature. Uptake was stopped by removing the transport medium and the monolayers were rapidly rinsed four times with 2 ml of ice-cold 109Cd-free medium to remove the excess radioactivity. In some experiments, these rinses were performed with stop solutions containing 2 mM EDTA to extract the external labile metal fraction from the cell surface (adsorption). Therefore, in the text, “accumulation” refers to the total amount of 109Cd measured in cell samples (adsorption plus uptake) whereas “uptake” data (transport) correspond to the EDTA-washed cells (accumulation minus adsorption). Cells were then solubilized in 1 N NaOH (0.5 ml) and aliquots of 0.3 ml were used for radioactivity determination with a gamma counter (Cobra II, Canberra, Packard Canada). Aliquots (50 ␮l) of the remaining suspension were kept for protein content determinations performed according to Bradford (1979), with bovine serum albumin used as the calibration standard. Experimental conditions for speciation studies. Experiments performed to evaluate the relative contribution of the free Cd 2⫹ ion to the total Cd cellular accumulation were designed according to speciation calculations. The total concentrations of Cd chlorocomplexes in the transport media (⌺[CdCl n2⫺n ]) were modified without changing [Cd 2⫹]. The chloride concentration was modified by substituting NaNO 3 for NaCl, to keep both [Na ⫹] and the ionic strength of the exposure media constant; Cd complexation by NO 3⫺ is negligible compared to by Cl ⫺. The total concentration of 109Cd added to the media, [ 109Cd] tot, had to be decreased along with [Cl ⫺], such that only [CdCl n2⫺n ] species were varied while the free [Cd 2⫹] was maintained constant at 43 nM (Table 1). Note that total [Ca] was also slightly corrected ([Ca] T was increased from 2.4 to 2.5 mM as [Cl ⫺] was raised from 9.7 to 147 mM) to ensure that the free [Ca 2⫹] levels were similar relative to those in the standard transport medium (2.1 mM). Because of chloride additions related to KCl (4.7 mM) and CaCl 2 (2.5 mM), still present in the transport media, the minimum [Cl ⫺] tested was 9.7 mM. The influence of Cd complexation by GSH (␥-Glu-Cys-Gly) or hmPC 3 (␥-Glu-Cys) 3-Ser on cellular accumulation was first investigated by adding increasing concentrations of the respective ligand (GSH, 50 to 250 ␮M; hmPC 3, 0.5 to 2.5 ␮M) to the standard transport medium containing 0.3 ␮M total 109Cd in each case. In these experiments, the total concentration of

31

CD UPTAKE AS A FUNCTION OF METAL SPECIATION

TABLE 1 Cadmium Speciation as a Function of Increasing Chloride Concentration Experiment no.

[ 109Cd] tot (nM)

[Cl ⫺] (mM)

[NO 3⫺] (mM)

[ 109Cd 2⫹] (nM)

[ 109CdCl ⫹] (nM)

[ 109CdCl 2] (nM)

[ 109CdCl 3⫺] (nM)

[ 109CdNO 3⫹] (nM)

[ 109CdHCO 3⫺] (nM)

1 2 3 4 5 6 7 8 9 10

70 90 110 133 157 182 208 236 265 300

9.7 25 40 55 70 85 100 115 130 147

137 122 107 92 77 62 47 32 17 0

43 43 43 43 43 43 43 43 43 43

12 31 49 68 87 106 124 143 161 182

— 1.7 4.4 8.4 14 20 28 37 47 60

— — — — — — — 2.6 3.6 5.1

6.4 5.7 5.0 4.3 3.6 2.9 2.1 — — —

6.3 6.3 6.3 6.3 6.3 6.4 6.2 6.4 6.4 6.3

Note. Cadmium speciation was calculated using MINEQL ⫹ chemical equilibrium program as described in the text. Components listed are those for which the concentrations were modified to ensure that [Cd 2⫹] remained constant as the chlorocomplex levels were modified. Species listed are those representing more than 1% of the total dissolved metal.

inorganic species [ 109Cd] inorg decreased with increasing complexation by GSH or hmPC 3 (Table 2). Another series of experiments was then designed to investigate specifically the accumulation of the respective complexes Cd–GSH and Cd– hmPC 3. Increasing concentrations of ligand (GSH, 13 to 820 ␮M; hmPC 3, 0.10 to 9.7 ␮M) were added to the standard transport medium, and total [ 109Cd] was adjusted in accordance with Cd affinity for each of the ligands in such a way that only [ 109Cd–GSH] and [ 109Cd– hmPC 3] were modified, i.e., [Cd 2⫹] was kept constant as in the experiments with variable [Cl ⫺] described above. Under these conditions, the relative concentrations of each of the major inorganic species (Cd 2⫹, CdCl ⫹, CdCl 2, and CdCl 3⫺ ) were unmodified. For all experiments, the preexposure washing step as well as the uptake stop after exposure were always performed in the absence of any organic ligand. Moreover, in accordance with our previous work showing that Cd uptake is linear over 3-min exposure times (Jumarie et al., 1999), the effect of Cd speciation on metal accumulation was tested using 3-min incubations under initial rate conditions. Speciation calculation. Cadmium speciation in each of the transport media used in this study was calculated using the MINEQL ⫹ chemical equilibrium program (Schecher and McAvoy, 1994) and the NIST stability constants database (Martell et al., 1998). The formation constants (log K) for Cd complexation in the standard transport medium (corrected to zero ionic strength) were CdCl ⫹, 1.98; CdCl 2, 2.6; CdCl 3⫺ 2.39; CdSO 4, 2.46; Cd(SO 4) 2⫺2 3.5; CdHCO 3⫹ 13.23; Cd(CO 3) 2⫺2 7.25; and Cd(CO 3) 3⫺4 6.22. For Cd(NO 3) n2⫺n species, formation constants were CdNO 3⫹ 0.5 and Cd(NO 3) 2 0.2. The conditional formation constants for Cd complexation by GSH and hmPC 3 were determined experimentally using an ion-exchange technique (Fortin and

Campbell, 1998) and equilibrium conditions similar to those in the transport media used for the uptake measurements (ionic strength of 0.2 M, pH 7.3). For these conditions, conditional log K values of 5.0 ⫾ 0.1 and 7.3 ⫾ 0.3 were determined for Cd complexation by GSH and hmPC 3, respectively, assuming a 1:1 complexation ratio. These values were corrected to zero ionic strength with the Davies equation before they were introduced into the MINEQL ⫹ database. Data Analyses All linear and nonlinear regression analyses of the accumulation/uptake data were performed using Enzfitter software (Robin J. Leatherbarrow, Copyright 1987) and the robust weighting routine. The errors associated with kinetic parameter values given in the text represent the standard errors of regression (SER). Materials All culture ware (Sarsted) was obtained from VWR Scientific (Toronto, Ontario, Canada). DMEM, penicillin, streptomycin, and trypsin were purchased from Gibco BRL (Grand Island, NY). FBS was obtained from Medicorp Inc. (Montreal, Quebec, Canada). The nonessential amino acids L-alanine, L-asparagine, L-glutamic acid, L-proline, and L-serine were obtained from Gibco BRL, whereas L-aspartic acid and glycine were purchased from Sigma Chemical Co. (St. Louis, MO). Labeled 109CdCl 2 was obtained from Dupont Canada Inc. (Mississauga, Ontario, Canada). Reduced GSH (␥-Glu-Cys-Gly) was purchased from Sigma Chemical Co., whereas hmPC 3 (␥-Glu-Cys) 3-Ser

TABLE 2 Cadmium Speciation as a Function of Increasing Concentrations of Organic Ligands [GSH] (␮M)

[ 109CdCl ⫹] (nM)

[ 109CdCl 2] (nM)

[ 109Cd 2⫹] (nM)

[ 109Cd–GSH] (nM)

[hmPC 3] (␮M)

[ 109CdCl ⫹] (nM)

[ 109CdCl 2] (nM)

[ 109Cd 2⫹] (nM)

[ 109Cd–hmPC 3] (nM)

50 100 250

100 61 30

33 20 9.9

23 14 6.9

189 231 282

0.5 1.0 2.5

90 45 19

30 15 6.3

21 10.3 4.4

205 258 319

Note. Calculated cadmium speciation in the presence of GSH and hmPC 3 concentrations ranging from 50 to 250 ␮M and from 0.5 to 2.5 ␮M, respectively. A fixed total 109Cd level of 0.3 ␮M was used in all cases. Ligands were added to the standard transport medium, the composition of which is described in the text.

32

JUMARIE ET AL.

was a special order synthesized at the Sheldon Biotechnology Center (McGill University, Montreal, Quebec, Canada). All salts and chemicals used for buffer preparation were ACS reagent grade or of higher purity.

RESULTS

Specific Uptake of Cd and Its Chlorocomplexes Equilibrium calculations indicated that Cd in the transport medium used was present mainly as five different inorganic species: 61% as CdCl ⫹, 20% as CdCl 2, 14% as Cd 2⫹, 2% as CdHCO 3⫹ and 2% as CdCl 3⫺ (for those species representing more than 1% of the total dissolved metal). The relative contributions of CdCl n2⫺n species to the total uptake of Cd was determined by measuring the initial 3-min accumulation of 109 Cd in 21-day-old TC7 monolayers, as a function of increasing [Cl ⫺] but at fixed free [ 109Cd 2⫹]. Results of this experiment are shown in Fig. 1 (circles), where the data points obtained for [Cl ⫺] ranging from 9.7 to 147 mM correspond to the experimental conditions 1 to 10 described in Table 1 (condition 10 corresponds to the standard experimental conditions). Though [ 109Cd 2⫹] was unchanged, a linear increase in 109Cd accumulation was clearly observed as a function of increasing [Cl ⫺], strongly suggesting that 109CdCl n2⫺n species contribute to Cd accumulation (Fig. 1A). The nonzero intercept accumulation value of 1.33 ⫾ 0.05 pmol/3 min/mg protein, obtained at [Cl ⫺] ⫽ 0 by linear regression analysis of the data points, can be attributed to the free Cd 2⫹ ion, the concentration of which was maintained at 43 nM for all experimental conditions (experiments 1 to 10). Equilibrium calculations indicate that virtually all of the dissolved metal would be present as Cd 2⫹ in the absence of any Cl ⫺ in the exposure medium; slight complexation between Cd 2⫹ and NO 3⫺ occurs at very low [Cl ⫺] ([CdNO 3⫹] ⫽ 6.4 nM for experiment 1). From the constant accumulation value related to [ 109Cd 2⫹], one can estimate that the aquo-ion contributes 25% of the total accumulation measured under standard conditions (experiment 10, 5.3 ⫾ 0.7 pmol/3 min/mg protein). This constant value of accumulation related to the Cd 2⫹ species can be subtracted from the experimental data to represent 109Cd accumulation as a function of increasing [ 109CdCl n2⫺n ] (Fig. 1B). The resulting data could be fitted successfully to Eq. (1) A 3 ⫽ k D 关S兴

(1)

in which A 3 is the initial 3-min accumulation, [S] is the concentration of chlorocomplexes ([ 109CdCl n2⫺n ]), and k D is the proportionality constant for the accumulation process with a value of 13.6 ⫾ 0.8 pmol/3 min/mg protein/␮M. As expected, the use of a stop solution containing 2 mM EDTA slightly decreased accumulation values (triangles). Under standard exposure conditions (experiment 10), 26% of the total accumulation can be related to 109Cd adsorption onto the

FIG. 1. Short-term (3-min) accumulation (circles) and uptake (triangles) of 109Cd in 21-day-old TC7 cells measured as a function of increasing concentrations of (A) Cl ⫺ or (B) 109CdCl n2⫺n . The data in B have been corrected for the estimated contribution of 109Cd 2⫹ to the total accumulation/uptake (nonzero intercept value of 1.33 ⫾ 0.06 pmol/3 min/mg protein in A). Experimental conditions corresponding to each data point are described in Table 1 (experiments 1 to 10). In all cases the free Cd 2⫹ ion concentration was maintained at 43 nM. Values shown are means ⫾ SEM evaluated on four different cell passages. The lines shown are the best-fit curves through the data points as obtained according to Eq. (1): k D ⫽ 13.6 ⫾ 0.8 and 11.9 ⫾ 2.8 pmol/3 min/mg protein/␮M for tracer accumulation and uptake as a function of increasing chlorocomplex concentrations, respectively.

cell surface; an uptake rate of 3.9 ⫾ 0.3 pmol/3 min/mg protein was measured using EDTA-containing stop solutions compared to 5.39 ⫾ 0.7 pmol/3 min/mg protein with the standard stop procedure. Considering that the EDTA-insensitive fraction of Cd accumulation represents metal uptake, it can thus be estimated that 34% of total uptake (accumulation corrected for adsorption) can be assigned to Cd 2⫹ under standard exposure conditions. For experiments with an EDTA stop solution, an apparent k D value of 11.9 ⫾ 0.3 pmol/3 min/mg protein/␮M was obtained for 109Cd uptake as a function of increasing [ 109CdCl n2⫺n ], an ⬃13% decrease in the proportionality constant of accumulation.

33

CD UPTAKE AS A FUNCTION OF METAL SPECIATION

would be observed if only inorganic species were taken up. To this end, 109Cd speciation in the presence of GSH or hmPC 3 was computed (Table 2) and the theoretical uptake values exclusively related to 109Cd inorg still present in the uptake medium were calculated (Fig. 2, dashed columns). Comparison between these calculated values and the experimental data clearly shows that the decrease observed in 109Cd accumulation does not correspond to prediction; in the presence of either of the two ligands, accumulation values were much higher than expected on the basis that only inorganic species are taken up. For example, although the percentage of total 109Cd present as inorganic species fell to 16% in the presence of 250 ␮M GSH, tracer accumulation only decreased to 46% of its control value. The discrepancies between predicted and observed values are even more pronounced with hmPC 3, since no significant differences could be observed in tracer accumulation even though the relative concentration of 109Cd inorg decreased by 90% in the presence of 2.5 ␮M hmPC 3. Specific Accumulation of Complexes

FIG. 2. Short-term (3-min) 109Cd accumulation in 21-day-old TC7 cells exposed to 0.3 ␮M 109Cd (filled columns) in the presence of increasing concentrations of GSH (A) or hmPC 3 (B). Values shown are means ⫾ SEM evaluated on three different cell passages. The decrease in total concentration of inorganic metal, resulting from increasing 109Cd complexation, has been estimated from the data described in Table 2; the theoretical corresponding tracer accumulation related to these residual levels of inorganic 109Cd has been estimated (dashed columns).

Cd Accumulation as a Function of Increasing Concentrations of Organic Ligands The influence of complexation with organic ligands on the short-term accumulation of Cd was first tested using coexposures to 0.3 ␮M 109Cd and increasing concentrations of GSH or hmPC 3. Results of this study clearly shows that 109Cd accumulation decreases with increasing concentrations of GSH added to the exposure media (Fig. 2A, filled columns). Quantitatively, these results are as expected, since a fixed total concentration of tracer was used ([ 109Cd] total ⫽ 0.3 ␮M) and enhanced complexation with the added organic ligand necessarily leads to lower levels of inorganic 109Cd species ([ 109Cd] inorg). In contrast to GSH, the presence of hmPC 3 had no significant effect on 109 Cd uptake. However, in both cases, the results obtained may be compared to the theoretical decrease in accumulation that

109

Cd–GSH and

109

Cd– hmPC 3

The specific accumulation of Cd–GSH and Cd– hmPC 3 was further studied by measuring the 3-min accumulation of 109Cd as a function of increasing Cd complexation by GSH or hmPC 3, but at fixed total [ 109Cd] inorg. The results of these studies are presented in Figs. 3 (GSH) and 4 (hmPC 3), where the data points correspond to the experimental conditions described in Tables 3 and 4, respectively (experiment 1 corresponds to the standard conditions under which Cd speciation is described in Table 1 for 147 mM Cl ⫺). Note that, in all these experiments, and in contrast to the previous one, [ 109Cd] was increased in concert with the increase in [GSH] or [hmPC 3], such that [ 109Cd 2⫹] and [ 109Cd] inorg were maintained at constant levels equal to those in the standard transport medium. Tracer accumulation increased linearly with GSH concentrations ranging from 13 to 245 ␮M and then tended to plateau to a maximal value of 13.23 ⫾ 1.60 pmol/3 min/mg protein (Fig. 3A). After correction of the data points for the accumulation value exclusively related to 0.3 ␮M Cd inorg (4.29 ⫾ 0.47 pmol/3 min/mg protein for experiment 1 in Fig. 3A), tracer accumulation values could be successfully fitted to the Michaelis–Menten Eq. (2) A3 ⫽

V max关S兴 K m ⫹ 关S兴

(2)

where [S] stands for 109Cd–GSH concentration and for which V max ⫽ 14.5 ⫾ 1.8 pmol/3 min/mg protein and K m ⫽ 3.7 ⫾ 0.6 ␮M (Fig. 3B). On the other hand, no plateau could be observed in the presence of hmPC 3 concentrations ranging from 0.1 to 9.7 ␮M (Fig. 4A); the data corrected for tracer accumulation related to inorganic Cd species (5.3 ⫾ 0.7 pmol/3 min/mg protein for Exp. 1 in Fig. 4A) could be successfully analyzed

34

JUMARIE ET AL.

TABLE 3 Cadmium Complexation with GSH Experiment no.

[ 109Cd] tot (␮M)

[GSH] (␮M)

[ 109Cd–GSH] (␮M)

1 2 3 4 5 6 7 8 9 10 11 12

0.3 0.4 0.8 1.2 1.6 2.0 2.4 2.8 3.5 4.5 5.0 6.0

0 13 71 129 187 245 300 360 460 600 680 820

0 0.1 0.5 0.9 1.3 1.7 2.1 2.5 3.2 4.2 4.7 5.7

Note. Increasing 109Cd complexation with GSH was calculated using a log K value of 5.0 as explained in the text. The concentration of each of the main inorganic species remained unchanged relative to standard exposure conditions in the presence of 147 mM Cl ⫺ and the inorganic Cd speciation is as listed in Table 1.

FIG. 3. Short-term (3-min) 109Cd accumulation in 21-day-old TC7 cells measured as a function of increasing GSH (A) or 109Cd–GSH (B) concentrations after correction of the data for the contribution of the inorganic metal to the total accumulation (4.29 ⫾ 0.47 pmol/3 min/mg protein). Experimental conditions corresponding to each data point obtained are described in Table 3 (experiments 1 to 12). In all cases the respective levels of each of the inorganic species were maintained unmodified relative to the standard exposure conditions described in Table 1 (experiment 10). Values shown are means ⫾ SEM evaluated on five determinations on the same subculture. The line shown in B is the best-fit curve through the data points according to Eq. (2): K m ⫽ 3.73 ⫾ 0.64 ␮M; V max ⫽ 14.47 ⫾ 1.78 pmol/3 min/mg protein.

109

according to Eq. (1), in which [S] now stands for [ Cd– hmPC 3], with an apparent k D value of 8.8 ⫾ 0.2 pmol/3 min/mg protein/␮M (Fig. 4B). Since accumulation of 109 Cd–GSH clearly began to saturate at complex concentrations higher than 3 ␮M (Fig. 3B), direct comparisons of metal accumulation with respect to metal speciation were performed for lower Cd concentrations. Equilibrium calculations were used to define the assay conditions such that equal concentrations (0.3 ␮M) of 109 Cd inorg , 109 Cd—GSH, and 109 Cd– hmPC 3 were compared. Figure 5A shows accumulation (filled columns) and uptake (dashed columns) data for 0.3 ␮M 109 Cd inorg alone or in the presence of equimolar concentrations of either 109 Cd–GSH or 109 Cd– hmPC 3 ; Fig. 5B shows these data obtained in the

presence of the organic complexes after correction for the inorganic contribution to accumulation/uptake. Both 109 Cd– GSH and 109 Cd– hmPC 3 appeared to be transported; uptake data of 2.0 ⫾ 0.8 and 3.1 ⫾ 1.0 pmol/3 min/mg protein were obtained, respectively, using the EDTA washing procedure. Compared to the measured uptake of 4.2 ⫾ 0.6 pmol/3 min/mg protein for 109 Cd inorg , it can be estimated that transport for equimolar concentrations of Cd–GSH and Cd– hmPC 3 would be 0.48 and 0.73, as efficient as transport of TABLE 4 Cadmium Complexation with hmPC 3 Experiment no.

[ 109Cd] tot (␮M)

[HmPC 3] (␮M)

[ 109Cd–HmPC 3] (␮M)

1 2 3 4 5 6 7 8 9 10 11 12

0.3 0.4 0.8 1.2 1.7 2.0 2.4 2.8 3.5 4.5 5.1 6.0

0 0.1 0.9 1.5 2.3 2.9 3.5 4.3 5.5 7.1 8.1 9.7

0 0.06 0.5 0.9 1.4 1.7 2.1 2.5 3.2 4.2 4.8 5.7

Note. Increasing 109Cd complexation with hmPC 3 was calculated using a log K value of 7.3 as explained in the text. The concentration of each of the main inorganic species remained unchanged relative to standard exposure conditions in the presence of 147 mM Cl ⫺ and the inorganic Cd speciation is as listed in Table 1.

CD UPTAKE AS A FUNCTION OF METAL SPECIATION

35

DISCUSSION

Role of Cadmium Chlorocomplexes in Cd Uptake In the present study, we have used equilibrium calculations to design experiments to discriminate between the free Cd 2⫹ ion and the suite of chlorocomplexes (CdCl n2⫺n ), and thus to evaluate i) if Cd uptake is exclusively related to the concentration of Cd 2⫹, and, if not; ii) what is the relative contribution of this species to the total cellular accumulation of Cd. Our data obtained under experimental conditions where [ 109Cd 2⫹] remained unchanged clearly show that tracer accumulation increases linearly as a function of increasing [ 109CdCl n2⫺n ] (Fig. 1), being 4-fold higher in the presence of 147 mM Cl ⫺ than in the low-chloride medium, where Cd is present almost entirely as Cd 2⫹. To our knowledge, these results are the first evidence that chlorocomplexes contribute to Cd transport into intestinal cells; metal uptake is not exclusively correlated to the free Cd 2⫹ concentration in the exposure medium. Similar conclu-

FIG. 4. Short-term (3-min) 109Cd accumulation in 21-day-old TC7 cells measured as a function of increasing hmPC 3 (A) or 109Cd– hmPC 3 (B) concentrations after correction of the data for the contribution of the inorganic metal to the total accumulation (5.31 ⫾ 0.68 pmol/3 min/mg protein). Experimental conditions corresponding to each data point obtained are described in Table 4 (experiments 1 to 12). In all cases the respective levels of each of the inorganic species were maintained unmodified relative to the standard exposure conditions described in Table 1 (experiment 10). Values shown are means ⫾ SEM evaluated on three different cell passages. The lines shown in B are the best-fit curves through the data points according to Eq. (1): k D ⫽ 8.77 ⫾ 0.20 pmol/3 min/mg protein/␮M.

Cd inorg at low metal concentrations, where the membrane transport system is not saturated. Note that, in agreement with previous results (Fig. 1), 20% of total accumulation was related to metal adsorption under standard conditions. Also, the contribution of adsorption to the total accumulation of tracer did not increase in the presence of organic ligands, suggesting that Cd–GSH and Cd– hmPC 3 are not significantly adsorbed compared to the inorganic species of Cd (Fig. 5A). This hypothesis is supported by Fig. 5B, showing no significant differences between the accumulation and uptake data obtained for the organic metal complexes alone following correction for the contribution of 0.3 ␮M Cd inorg .

FIG. 5. (A) Short-term (3-min) accumulation (filled columns) and uptake (crossed columns) of 0.3 ␮M inorganic 109Cd in the presence of equimolar concentrations of 109Cd–GSH or 109Cd– hmPC 3. (B) Data points obtained under conditions of organic complexation have been corrected for the inorganic contribution to the total accumulation/uptake. Values shown are means ⫾ SEM evaluated on five determinations on the same subculture.

36

JUMARIE ET AL.

sions have recently been reached concerning Cd accumulation in rooted plants: Cd uptake by Swiss chard increased 1.3-fold as [Cl ⫺] increased to 120 mM in EDTA-containing solutions used to buffer the free [Cd 2⫹] at a constant level (analogous to our Fig. 1) (Smolders and McLaughlin, 1996). These results and ours strongly suggest that cells as different as plant and mammalian cells do transport Cd as chlorocomplexes, possibly via a charge-neutral cotransport mechanism. However, although a number of studies have suggested the involvement of anion exchangers in the uptake of trace metals (Alda and Garay, 1990; Kalfakakou and Simons, 1990; Lou et al., 1991), the involvement of such antiports is much less probable. Lou et al. (1991) reported a 4,4⬘-diisothiocyanostilbene-2,2⬘-disulfonic acid-sensitive component for Cd absorption by human red blood cells that is stimulated by increasing levels of external HCO 3⫺ in the presence of an excess of Na ⫹. The authors concluded that the Cl ⫺–HCO 3⫺ exchanger was involved in Cd uptake, without, however, any evidence for a need to restore pH i. They speculated about the possible role of chlorocomplexes in metal absorption, but they did not report any speciation results for their exposure media. More Efficient Uptake for Cd 2⫹ than for Chlorocomplexes In interpreting data obtained from experiments designed to correlate rates of Cd uptake with particular metal species in solution, it is important to focus on the initial uptake rates and to ensure that the concentrations of the various metal species in solution are well below the levels at which the membrane transport systems become saturated. The use of saturating concentrations and lengthy exposure times may lead to misinterpretation of the data. Our previous studies performed under standard conditions (147 mM Cl ⫺) revealed the involvement of a specific system of transport with a K m value of about 3 ␮M (Jumarie et al., 1997, 1999); in the present studies we have worked at total Cd concentrations ranging from 0.07 to 0.3 ␮M, well below this K m value. Under such conditions, the Michaelis–Menten Eq. (2), where [S] may represent any of those inorganic metal species present in solution, reduces to the linear Eq. (1): when [S] ⬍⬍ K m , A 3 ⫽ k D [S] where k D ⫽ V max/K m . The observation for these conditions that 34% of total Cd uptake may be related to the free Cd 2⫹ aquo-ion, even though this metal species contributes only 14% of the total dissolved metal, allows some comparison of uptake efficiencies for Cd 2⫹ and CdCl n2⫺n . From the nonzero intercept uptake value, which can be attributed to the free Cd 2⫹ cation (43 nM; 1.33 ⫾ 0.05 pmol/3 min/mg protein in Fig. 1A), it can be estimated that, under conditions where membrane transport systems are not saturated, the short-term uptake of Cd 2⫹ would increase linearly with a proportionality constant of about 30.9 ⫾ 1.2 pmol/3 min/mg protein/␮M. Comparison of this value with that of 11.9 ⫾ 0.3 pmol/3 min/mg protein/␮M obtained for CdCl n2⫺n uptake (Fig. 1B) suggests that chlorocomplexes would be transported about 0.4 times as rapidly as

the free aquo-ion. Ideally, high Cd and chloride concentrations would have been used to characterize the affinity of the transport system specific to the chlorocomplexes; however, the resulting dramatic increase in the osmolarity of the exposure medium leads to cell shrinkage and the activation of cell volume regulatory processes, which compromise the interpretation of the experiments. Different Efficiencies of Accumulation for Cd–GSH and Cd– hmPC 3 Numerous studies have shown that complexation with organic ligands reduces the intestinal absorption of Cd in vivo as well as in vitro (Kojima et al., 1985; Fox, 1988; Morberg Wing, 1993). It is generally assumed that Cd bound to organic ligands is not taken up by intestinal cells, and thus that metal complexation by such ligands will reduce metal bioavailability for absorption. However, none of these studies have demonstrated that the resulting decrease in metal absorption was quantitatively related to modifications in metal speciation. Our results obtained with GSH and hmPC 3 confirm that Cd accumulation by intestinal cells may decrease upon ligand addition to the exposure media but much less than predicted on the basis the free-ion model (Fig. 2), strongly suggesting that organic complexes are also taken up. However, these initial observations remain ambiguous with respect to the accumulation of Cd–L complexes, since the concentrations of each of the inorganic metal species are modified as the ligand concentration [L] increases. Therefore, an experimental procedure was designed to study the specific uptake of Cd–GSH and Cd– hmPC 3 under conditions where inorganic metal speciation remained unmodified. Kinetic studies of Cd–GSH accumulation clearly revealed the involvement of a specific component of accumulation with a K m of 3.7 ⫾ 0.6 ␮M and V max of 14.5 ⫾ 1.8 pmol/3 min/mg protein. These kinetic parameter values predict an accumulation of about 1.1 pmol Cd–GSH/mg protein following a 3-min exposure to 0.3 ␮M Cd–GSH. For comparison, kinetic parameter values obtained previously (Jumarie et al., 1997) for inorganic Cd following a 1-min exposure (K m ⬃ 3 ␮M; V max ⬃ 15 pmol/min/mg protein and k D ⬃ 0.9 pmol/min/mg protein/ ␮M) yield a cellular accumulation of about 4.9 pmol Cd inorg/mg protein following a 3-min exposure to 0.3 ␮M inorganic Cd. It follows that the accumulation ratio Cd–GSH/Cd inorg would be about 0.22 for nonsaturating initial rate conditions. Similarly, the proportionality constant of 8.8 ⫾ 0.2 pmol/3 min/mg protein/␮M determined for Cd– hmPC 3 accumulation (Fig. 4B) predicts an accumulation value of about 2.6 pmol CdhmPC 3/mg protein following exposure to 0.3 ␮M Cd– hmPC 3. The uptake ratio Cd– hmPC 3/Cd inorg would be equal to 0.53; Cd bound to the phytochelatin hmPC 3 is accumulated half as fast as the inorganic metal. These estimates compare quite well with the results presented in Fig. 5B showing the relative uptake of 0.3 ␮M Cd inorg, Cd—GSH, and Cd– hmPC 3.

CD UPTAKE AS A FUNCTION OF METAL SPECIATION

The Involvement of a Specific System for Cd–GSH Uptake Molecular mimicry is of primary interest for heavy metal toxicity and it has been suggested that methylmercury compounds may be transported across biological membranes via transport systems specific to amino acids or GSH-S conjugates (Vincenzini et al., 1992; Clarkson, 1993). This raises the question as to whether or not Cd–GSH and Cd– hmPC 3 are transported via specific transport systems designed for GSH and hmPC 3, respectively. Note that, although no saturation could be observed under our experimental conditions for Cd– hmPC 3 up to a concentration 6 ␮M, the involvement of a saturable component for the uptake of this compound cannot be excluded; a K m value far higher than 6 ␮M is conceivable. One may also ask if Cd–GSH is transported as an intact entity or does the metal tripeptide complex undergo hydrolysis before being transported. Studies on the intestinal absorption of methylmercury suggest that some of the complexes formed with GSH could be transported via a ␥-GT-independent system (Urano et al., 1990). Such a transport system has been characterized before: a K m of 23 ⫾ 3 ␮M and a V max of 120 pmol/mg protein/10 s were determined for intact GSH in intestinal brush-border membrane vesicles (Vincenzini et al., 1989). Even if we consider a general enrichment factor of 10 for brush-border membrane preparations, this V max value is clearly much higher than the maximal velocity determined in the present study for the accumulation of Cd–GSH by Caco-2 cells. However, the K m value reported by Vincenzini et al. (1989) and the value estimated in the present study (3.7 ⫾ 0.6 ␮M) agree in that they both suggest that a very high affinity transport system is involved. The role of phytochelatins in Cd retention in contaminated plants has been well characterized (Grill et al., 1985; Bernhard and Ka¨gi, 1987) and the intestinal absorption of Cd bound to PCs is a relevant toxicological endpoint. Using an ion-exchange technique developed in our laboratories, we obtained conditional log K values of 5.0 ⫾ 0.1 and 7.3 ⫾ 0.3 for Cd complexation with GSH and hmPC 3, respectively, showing that Cd– hmPC 3 complexes are at least 100-fold stronger than the Cd–GSH complex. To our knowledge, there are no data available on the intestinal absorption of PCs per se. However, it has been shown in vivo that Cd bound to PC was less absorbed by rats than was Cd administered as an aqueous CdCl 2 solution (Fujita et al., 1993), a result that agrees with our present in vitro results. Though PCs are GSH-related plant peptides, in contrast to GSH, it is unlikely that they could be absorbed intact; the hmPC 3, (␥-Glu-Cys) 3-Ser, used in our study seems too large to be transported across biological membranes as an intact entity. Elucidation of the metabolic reactions involved in Cd– hmPC 3 absorption is a complex question; it will require a complete characterization of the specific transport system as well as estimation of the relative contribution of several possible systems. Nevertheless, our results clearly

37

show that such studies could be undertaken with the human enterocytic-like cells Caco-2. In conclusion, we have shown that Cd uptake by intestinal cells exposed to dissolved metal in an isotonic sodium chloride solution is not directly correlated with Cd 2⫹ levels; it is suggested that chlorocomplexes are taken up, though less rapidly than the free aquo-ion. Cadmium is expected to be bound to thiol peptides in animal- as well as plant-derived food products. To our knowledge, there are no reports concerning the stability of either Cd–GSH or Cd–PCs in the presence of gastric or intestinal enzymes. Some studies have shown that the stability of Cd–metallothionein complexes (Cd–MT) in artificial human gastric juice decreased with acidic pH; above pH 3.5 nearly all the Cd was resistant toward proteolysis by trypsin (Klein et al., 1986). However, for lower pH values (2.5 and 1.7) most of the protein was digested, though some Cd-containing peptides with 25–30 amino acids similar to the ␣-domain of the MT were still present. Another study suggested that a significant level of the Cd–MT complex initially present in pig kidney would survive both cooking and gastrointestinal digestion (Crews et al., 1989). It is difficult to extrapolate from the results obtained with MT to predict the stability of Cd– GSH or Cd–PCs following ingestion. However, the previous reported results do suggest that a significant part of the original organic complexes would reach the intestinal lumen intact; our present study suggest that these residual metal complexes may contribute to Cd assimilation. Ideally, the levels of complexes that would survive gastric digestion in vivo should be taken into account when evaluating their oral bioavailability. Binding of the metal to the peptide would presumably limit direct metal interactions with intracellular thiols critical for the cell survival. However, metal exchange between the absorbed ligand and other intracellular metabolites cannot be excluded once the metal is in the cell. ACKNOWLEDGMENTS This research was supported by a grant from the Canadian Network of Toxicology Centers/Environment Canada (Green Plan)

REFERENCES Alda, J. O., and Garay, R. (1990). Chloride (or bicarbonate)-dependent copper uptake through the anion exchanger in human red blood cells. Am. J. Physiol. 259, C570 –C576. Berglund, M., Akesson, A., Nermell, B., and Vahter, M. (1994). Intestinal absorption of dietary cadmium in women depends on body iron stores and fiber intake. Environ. Health Perspect. 102, 1058 –1066. Bergmann, L., and Rennenberg, H. (1993). Glutathione metabolism in plants. In Sulfur Nutrition and Sulfur Assimilation in Higher Plants (L. J. De Kok, I. Stulen, H. Rennenberg, C. Brunold, and W. E. Rauser, Eds.), pp.109 –123. SPB Academic Publishing, The Hague. Bernhard, W. R., and Ka¨gi, J. H. R. (1987). Purification and characterization of atypical cadmium-binding polypeptides from Zea mays. Experientia Suppl. 52, 309 –315. Blais, A., Lecoeur, S., Milhaud, G., Tome´, D., and Kolf-Clauw, M. (1999).

38

JUMARIE ET AL.

Cadmium uptake and transepithelial transport in control and long-term exposed Caco-2 cells: The role of metallothionein. Toxicol. Appl. Pharmacol. 160, 76 – 85. Bradford, M. M. (1979). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248 –254. Campbell, P. G. C. (1995). Interactions between trace metals and aquatic organisms: A critique of the free-ion activity model. In Metal Speciation and Bioavailability in Aquatic Systems (A. Tessier and D. R. Turner, Eds.), pp. 45–102. Wiley, New York. Chantret, I., Rodolosse, A., Barbat, A., Dussaulx, E., Brot-Laroche, E., Zweibaum, A., and Rousset, M. (1994). Differential expression of sucraseisomaltase in clones isolated from early and late passage of the cell line Caco-2: Evidence for glucose-dependent negative regulation. J. Cell Sci. 107, 213–225. Clarkson, T. W. (1993). Molecular and ionic mimicry of toxic metals. Annu. Rev. Pharmacol. Toxicol. 32, 545–571. Crews, H. M., Dean, J. R., Ebdon, L., and Massey, R. C. (1989). Application of high-performance liquid chromatography inductively coupled plasma mass spectrometry to the investigation of cadmium speciation in pig kidney following cooking and in vitro gastro-intestinal digestion. Analyst 114, 895– 899. Fortin, C., and Campbell, P. G. C. (1998). An ion-exchange technique for free-metal ion measurements (Cd 2⫹, Zn 2⫹): Application to complex aqueous media. Int. J. Environ. Anal. Chem. 72, 171–194. Fox, S. (1988). Nutritional factors that may influence bioavailability of cadmium. J. Environ. Qual. 2, 175–179. Fujita, Y., El-Belbasi, H. I., Min, K. S., Onosaka, S., Okada, Y., Matsumoto, Y., Mutoh, N., and Tanaka, K. (1993). Fate of cadmium bound to phytochelatin in rats. Res. Commun. Chem. Pathol. Pharmacol. 82, 357–365. Grill, E., Winnacker, E. L., and Zenk, M. H. (1985). Phytochelatins: The principal heavy-metal binding peptides of higher plants. Science 230, 674 – 676. Hagen, T. M., Wierzbicka, G. T., Bowman, B. B., Aw, T. Y., and Jones, D. P. (1990). Fate of dietary glutathione: Disposition in the gastrointestinal tract. Am. J. Physiol. 259, G530 –G535. Iantomasi, T., Favilli, F., Marraccini, P., Magaldi, T., Bruni, P., and Vincenzini, M. T. (1997). Glutathione transport system in human small intestine epithelial cells. Biochim. Biophys. Acta 1330, 274 –283. Jumarie, C., Campbell, P. G. C., Berteloot, A., Houde, M., and Denizeau, F. (1997). Caco-2 cells line used as an in vitro model to study Cd accumulation in intestinal epithelial cells. J. Membr. Biol. 158, 31– 48. Jumarie, C., Campbell, P. G. C., Houde, M., and Denizeau, F. (1999). Evidence for an intracellular barrier to Cd transport through Caco-2 cell monolayers. J. Cell. Physiol. 180, 285–297. Kalfakakou, V., and Simons, T. J. (1990). Anionic mechanisms of zinc uptake across the human red cell membrane. J. Physiol. 421, 485– 497. Kang, Y. J. (1992). Exogenous glutathione decreases cellular cadmium uptake and toxicity. Drug Metab. Dispos. 20, 714 –718.

Klapheck, S., Chrost, B., Starke, J., and Zimmermann, H. (1992). ␥-Glutamylcysteinylserine–A new homologue of glutathione in plant of the family Poaceae. Bot. Acta 105, 174 –179. Klapheck, S., Fliegner, W., and Zimmer, I. (1994). Hydroxymethyl-phytochelatins [␥-(glutamylcysteine) n-serine] are metal-induced peptides of the Poaceae. Plant Physiol. 104, 1325–1332. Klein, D., Greim, H., and Summer, K. H. (1986). Stability of metallothionein in gastric juice. Toxicology 41, 121–129. Kojima, S., Kiyozumi, M., Mishima, M., Honda, T., and Naragawa, M. (1985). Effects of three proteins on absorption of cadmium in rats. Toxicology 34, 161–171. Kojima, Y. (1991). Definitions and nomenclature of metallothioneins. Methods Enzymol. 205, 8 –10. Lou, M., Garay, R., and Alda, J. O. (1991). Cadmium uptake through the anion exchanger in human red blood cells. J. Physiol. 443, 123–136. Martell A. E., Smith R. M., and Motekaitis R. J. (1998). NIST Critical Stability Constants of Metal Complexes Database, Version 5.0. U.S. Department of Commerce, Gaithersburg, MD. Meuwly, P., Thibault, P., and Rauser, W. E. (1993). ␥-Glutamyl-cysteinylglutamic acid–A new homologue of glutathione in maize seedlings exposed to cadmium. FEBS Lett. 336, 472– 476. Morberg Wing, A. (1993). The effects of whole wheat, wheat bran and zinc in the diet on the absorption and accumulation of cadmium in rats. Br. J. Nutr. 69, 199 –209. Morel, F. M. M. (1983). Principles of Aquatic Chemistry. Wiley-Interscience, New York. Murasugi, A., Wada, C., and Hayashi, Y. (1981). Purification and unique properties in UV and CD spectra of Cd-binding peptide 1 from Dchizosaccharomyces pompe. Biochem. Biophys. Res. Commun. 103, 1021–1028. Perrin, D. D., and Watt, A. E. (1971). Complex formation of zinc and cadmium with glutathione. Biochim. Biophys. Acta 230, 96 –104. Schecher, W. D., and McAvoy, D. (1994). MINEQL⫹: A Chemical Equilibrium Program for Personal Computers. Version 3.01. Environ. Research Software, Hallowell, ME. Smolders, E., and McLaughlin, M. J. (1996). Effect of Cl on Cd uptake by Swiss chard in nutrient solutions. Plant Soil 179, 57– 64. Tahvonen, R., and Kumpulainen, J. (1994). Lead and cadmium contents in Finnish breads. Food Addit. Contam. 11, 621– 631. Urano, T., Iwasaki, A., Himeno, S., Naganuma, A., and Imura, N. (1990). Absorption of methylmercury compounds from rat intestine. Toxicol. Lett. 50, 159 –164. Vincenzini, M. T., Iantomasi, T., and Favilli, F. (1989). Glutathione transport across intestinal brush-border membranes: Effect of ions, pH, ⌬⌿, and inhibitors. Biochim. Biophys. Acta 987, 29 –37. Vincenzini, M. T., Favilli, F., and Iantomasi, T. (1992). Intestinal uptake and transmembrane transport systems of intact GSH; characteristics and possible biological role. Biochim. Biophys. Acta 1113, 13–23.