ELSEVIER

Loss of the

Toxicology 120 (1997) 37-46

metal binding properties of metallothionein induced by hydrogen peroxide and free radicals

I. JimCnez a, M. Gotteland b, A. Zarzuelo c, R. Uauy a, H. Speisky ‘,*

a Unidad de Bioquimica Farmacoldgica, Instituto de Nutricidn y Tecnologia de 10s Alimentos, Vniversidad de Chile, Santiago, Chile b Vnidad de Gastroenterologia, Instituto de Nutrici6n y Tecnologia de 10s Alimentas, Vniversidad de Chile, Santiago, Chile

’ Depto. de Farmacologia, Facultad de Farmacia, Vniversidad de Granada, Granada, Spain

Received 21 October 1996; accepted 22 January 1997

Abstract

The relationship between the metal-binding properties of metallothionein (MT) and its ability to interact with peroxides and free radicals was explored in vitro. The binding of lo9Cd to MT and the thiol density of the protein were determined after incubation of a purified Zn/Cd-metallothionein preparation with either hydrogen peroxide alone, or with a number of free radical generating systems. Exposure of MT to H202, whether in the presence or absence of Fe* + , resulted in the progressive loss of the thiol residues of the protein and led to a parallel decrease of its lopCd-binding capacity. These changes correlated with r values of 0.999 (P = 0.001) and 0.998 (P = O.OOl), in the absence and presl:nce of iron, respectively. The effects of H,O,, alone or plus Fe* + , on MT were completely prevented by catalase, but totally unaffected by superoxide dismutase or desferrioxamine. Exposure of MT to xanthine/xanthine oxidase also led to thiol oxidation and to a concomitant loss of the Cd-binding properties. In this system, both changes correlated with an I of 0.993 (P = 0.001) and were completely inhibited by superoxide dismutase. Exposure of MT to the peroxyl radical generator, 2,2’-azobis(2-amidinopropane) dihydrochloride (AAPH), resulted in the progressive loss of its the metal-binding properties and its thiol residues, both changes correlating with an r of 0.986 (P = 0.002). The ability of MT to bind ‘OpCd, lost as a result of its prior exposure to either H202 alone, H,O, plus Fe* + , xanthine/xanthine oxidase, or to AAPH was, in all cases, completely recovered after incubation of the modified protl;in with dithiothreitol. These results indicate that H,O, alone, and/or the oxygen-derived species, superoxide anion and peroxyl radicals, can all directly interact in vitro with MT to modify the protein oxidatively, and suggest that, under in vivo conditions, these species may be implicated as modifying factors of the metal-binding capacity of metallothionein. 0 1997 Elsevier Science Ireland Ltd.

Keywords: Metallothionein; Thiol oxidation; Metal-binding; Free radicals; Hydrogen peroxide

Abbreviations: MT, metallothionein; AAPH, 2,2’-azobis(2-amidinopropane) dihydrochloride; X0, xanthine oxidase; Xan, xan- thine; DTT, dithiol:hreitol; Nbs,, 5,5’-dithiobis(2-nitrobenzoic acid); DESF, desferrioxamine; CAT, catalase; SOD, superoxide dismutase.

* Corresponding author, Instituto de Nutrici6n y Tecnologia de 10s Alimentos, Universidad de Chile, Casilla 138-11 Santiago, Chile. Tel: + 56 2 6781448; fax: + 56 2 2214030; e-mail: hspeisky @ uec.inta.uchile.cl.

0300-483X/97/$17.00 0 1997 Elsevier Science Ireland Ltd. All rights reserved.

PIZ SO300-483X(97)03638-X

38 I. Jimtner et al. /Toxicology 120 (1997) 37-46

1. Introduction

Metallothionein (MT) is a low molecular weight (6-7 kDa) protein which plays a major biological role in the cellular homeostasis of cop-per and zinc, possibly by affecting the transport and storage of

these two essential trace metals (Cousins, 1985; Bremner, 1987). In most animal tissues, MT has a high metal content, occurring with either 7 gram atoms of zinc or with 12 gram atoms of copper per mole (Kagi et al., 1990). Metallothionein is primar- ily characterized by its high thiol content with 20 out of its 61 amino acids being cysteine, all of which are in their reduced state and involved in the protein’s metal binding properties. The affinity of metals for the binding sites on MT is substantially greater for copper and cadmium ions than for zinc (Li et al., 1980). These metals are tightly bound to

the protein and can be removed by the exposure of MT to non-physiological acidic pH, yielding the apoprotein. Upon neutralizing the medium, metal- lothionein can be reconstituted as a metalloprotein (Li et al., 1980).

As a result of its metal binding properties, a considerable number of studies support a metal detoxifying role for MT. Using a large variety of in vitro and in vivo experimental models, inverse correlations have been established between the intracellular levels of metallothionein (from basal to induced) and the susceptibility of cells exposed to toxic doses of metals capable of binding MT (Goering and Klassen, 1984; Freedman and Peisach, 1989; Templeton and Cherian, 1991). While the concentration of such metals appears to be a major determinant of their acute cytotoxicity, induction of MT synthesis may also occur in response to the repeated exposure of cells to non-toxic doses of such metals (Lehman-McKee- man et al., 1988).

Although most of the biological functions as- cribed to metallothionein are related to its metal- binding properties, MT has recently received increasing attention as result of its ability to act as a free radical scavenger (Sato and Bremner, 1993). Early work by Thornalley and Vasak (1985) showed that purified Zn-MT can react very rapidly with hydroxyl radicals (HO. ). The purported abil- ity of MT to react with HO. has led to the

proposal that this protein could play both antiox- idant and cytoprotective roles against the toxicity induced by these and possibly other free radicals (Hidalgo et al., 1988; Sato and Bremner, 1993). Although the contention that MT plays a biologi- cal role as an antioxidant is still a debatable matter (Lazo and Pitt, 1995) a large number of observa- tions provide strong support for such a role. Purified Zn-MT has been shown to inhibit the lipid peroxidation of erythrocyte ghosts (Thomas et al., 1986) and to protect DNA against degradation induced by either a Fenton reaction (Abel and Ruiter, 1989) or by X-radiation (Greenstock et al., 1987). On the other hand, support for the claim of a role of MT in cytoprotection arises from various studies showing that interventions leading to the induction of MT in cultured cells, or in intact animals, render both the in vitro induced cells (Bakka et al., 1982; Coppen et al., 1988; Mello- Filho et al., 1988; Renan and Dowman, 1989; Leyshon-S&land et al., 1993) and the experimen- tal animals (Cagen and Klaassen, 1979; Satoh et al., 1988) substantially less susceptible or com- pletely resistant to the action of chemicals or physical conditions known to promote - via free radicals - peroxidation and cytotoxicity.

Although the exact mechanism(s) underlying the free radical scavenging properties of MT remain to be established, it has been suggested that its thiol residues may be the primary target for the reaction with HO. radicals (Thornalley and Vaslk, 1985). Yet, the thiol groups of MT could also be impli- cated in its antioxidant action through their role in the binding and subsequent thiol oxidation-depen- dent release of zinc atoms (Thomas et al., 1986). Zn-MT has been shown to protect and stabilize membranes undergoing lipid peroxidation, and Zn atoms released from the protein have been sug- gested to act by themselves as inhibitors of this process (Bray and Bettger, 1990). Furthermore, thiol clusters in MT have been shown to chelate iron in vitro preventing the metal from catalyzing Fenton-dependent oxidative reactions (Kennedy et al., 1993).

Considering that the thiol residues in the MT molecule appear closely linked with both its ability to interact with radical species and its capacity to bind metals ions, in the present work a possible relationship between these two properties of the

I. Jim.hez et al. /Toxicology 120 (1997) 37-46 39

protein was addressed. Purified Zn/Cd-MT was incubated with several free radical-generating sys- tems and the effects of such exposure were contin- uously monitored on both the thiol density and on the metal binding capacity of the protein. Since during the course of our work we observed that, at least in the case of the Fenton reaction, the delete- rious effects on MT were not prevented by the addition of the iron chelator desferrioxamine, we decided to also address the effects of hydrogen peroxide alone on the thiol density and metal binding properties of this protein.

2. Material and ,methods

2.1. Reagents

The following chemicals were purchased from Sigma, St. Louis MO, USA: allopurinol (A-8003), catalase (10 000--25 000 U/mg, C-40), chelex-100 (C-7901) 5,5’-dithiobis(2nitrobenzoic acid) (D- 8 130), bovine helmoglobin (H-2625), Zn/Cd-metal- lothionein from rabbit liver (M-7641), superoxide dismutase (30000 U/mg, S-7008) and xanthine oxidase (l-2 U/mg, X-2500). Desferrioxamine was provided as Desferal@ by Ciba-Geigy, Chile. Ra- dioactive isotope lo9Cd as CdCl* was from Du Pont Co., hydrogen peroxide as Perhydrol@ 30%, and FeSO,. 7H20 were purchased from Merck. 2,2’- Azobis(2-amidinopropane) dihydrochloride was obtained from Wako Pure Chemical Industries Ltd., Osaka, Japan, dithiothreitol was obtained from AMRESCO@, Solon, OH and CdCl, . Hz0 from Riedel-De Haen AG, Seelze-Hannover, Germany.

2.2. Free-radical generating systems

2.2.1. H202 _+ Fe2 + The incubation system consisted of MT (0.1

mg/ml), in the presence or absence of FeSO, (10 PM), in 100 mM potassium phosphate buffer, pH 7.4. All reagents were diluted in the buffer solu- tion pretreated with chelex-100. The reaction was initiated by the addition of H,O, to a final con- centration of 500 PM, and the system was contin- ually shaken (76 Hz) in a water bath at 37°C (results from these experiments are depicted in Fig. lA, Fig. 1B and Fig. 1C). In other indepen-

60-

0 10 2@ 30 40 50

so-

-20 J I 0 20 40 60 80

% DECREASE OF TEIOLS

Fig. 1. Time course of the effects of the exposure of metalloth- ionein to a H,O, plus Fe2 + mixture. Data are expressed as % of decrease of the “‘Cd-binding capacity (Fig. IA), and % of decrease of thiol content (Fig. 1B) resulting from incubating MT (0.1 mg/ml) in the presence of FeSO, (10 PM), in 100 mM potassium phosphate buffer, pH 7.4. The reaction was ini- tiated by the addition of H,O, to a final concentration of 500 PM. Each plotted point corresponds to the mean of at least three independent experiments (S.E.M. were always less than 5%). Assays were conducted as described in Section 2. Fig. 1C depicts the correlation between the % of decrease of the ro9Cd-binding capacity and its corresponding % of decrease of thiol MT content. The line drawn is based on a linear regres- sion analysis (r’ = 0.996).

40 I. Jinknez ef al. /Toxicology 120 (1997) 37-46

Table 1 Exposure of metallothionein to hydrogen peroxide, in the

absence/presence of iron, and to a xanthineixanthine oxidase

system: effects of superoxide dismutase, catalase and desfer-

rioxamine

Incubation sys-

terns

‘% Binding remaining ‘%I SH remaining

System A

MT 100

MT/H,O, 63”

MT/ H,O,/SOD 64

MT/ H,O,/DESF 67b

100

69”

70

66b

System B

MT/Fe’+ 100 100

MT/H,OJFe”+ 60” MT/HO/ Fe Zt

CA;: 2 / 1:: 100

MT/H,O,/ Fe2+/ 70b

DESF

64b

System C

MT/X0

MT/Xan/XO

MT/Xan/XO/

SOD

100 100 71” 79”

100 100

MT/Xan/XO/

CAT

78b 84b

Results are presented as the percent remaining of binding and

thiol content of MT (0.1 mg/ml) following its incubation at

37°C with H,O, (250 PM) during 30 min (system A), or with

HzO, (500 ,uM) plus FeSO, (IO PM) during 15 min (system

B), in 100 mM potassium phosphate buffer, pH 7.4. In system

C, MT (0.1 mg/ml) was incubated at 37°C during IO min in

the presence of xanthine oxidase (0.01 U/ml) and xanthine (1 .O

mM), in 100 mM potassium phosphate buffer, pH 7.4. The

metal composition of MT was: 6.7 mol% Cd and 0.5 mol%

Zn. When added, SOD, CAT and DESF were present at a

final concentration of 875 U/ml, 2500 U/ml and 500 PM,

respectively. The ‘09Cd-binding capacity and thiol content

were assayed as described in Section 2. Data corresponds to

the means of 2-3 experiments. The statistical comparison

between incubations a and b gave P-values greater than 0.05.

dent experiments the final concentration of hydro- gen peroxide was either 250 ,uM (without addition of iron; system A, Table 1) or to 500 PM (with 10 /LM Fe2+ added; system B, Table 1). Aliquots were removed at different times and placed in a centrifuge tube containing 2500 U/ml of catalase to stop the reaction. These samples were used for the Cd binding assay and for the determination of

the thiol density of MT (uide infra). Tubes taining iron but not H,O, were included as trols.

2.2.2. Xanthinelxanthine oxidase The incubation system consisted of 100

con- con-

mM

potassium phosphate buffer, pH 7.4, containing 0.1 mg/ml of MT and 0.01 U/ml of xanthine oxidase (X0). All chemical and reagent solutions were prepared in phosphate buffer previously treated with chelex-100. The incubations were conducted at 37°C in a water bath shaken at 76 Hz. The reaction was initiated by the addition of xanthine (Xan) to a final concentration of 1 .O mM (system C, Table 1). Aliquots of the incubation mixture were removed at different times and the reaction was stopped by their addition into Ep- pendorf tubes containing ice-cold 0.25 mM allop- urinol. These samples were used for the Cd binding assay and for the determination of the thiol density of MT. Tubes containing no xan- thine were included as controls.

2.2.3. AAPH-system Zero-point-one mg/ml of MT in 100 mM potas-

sium phosphate buffer pH 7.4 was incubated in a water-bath at 37°C under aerobic conditions. The reaction was initiated by the addition of 25 mM 2,2’-azobis(2-amidinopropane) dihydrochloride (AAPH) to the system. Both solutions were freshly prepared in chelex-loo-pretreated buffer. Aliquots were removed at different incubation times and were diluted (1:9) in ice-cold potassium phosphate buffer to stop the reaction. The diluted samples were used to assay the “‘Cd-binding capacity and thiol content of MT. In order to study the oxidative effect of AAPH on the ‘09Cd- binding capacity of MT, and the possible re- versibility of such modification by subsequent chemical reduction of the protein (see below), MT was separated from AAPH. The latter was needed since the residual concentration of AAPH in the incubated samples was found to interfere with the heat-dependent step involved in the Cd-binding assay. Thus, 1:9 diluted samples (containing ap- proximately 2 ,ug of MT) were filtered at 3900 x g during 100 min, at 4°C using an Ultrafree@-MC Millipore microcentrifuge filter (MW cut-off point

I. Jinknez et al. / Toxicology 120 (1997) 37-46 41

5000, Millipore Corp., Bedford, MA). To assure the removal of any AAPH remaining on the filter surface, 100 ~1 of 10 mM Tris buffer, pH 7.4, were added (onto the filter and the filtration was repeated. MT was quantitatively recovered by the addition of 2 x 150 ~1 portions of the same Tris buffer to the filter, followed by vigor- ous vortexing during 1 min. This whole proce- dure systematically allowed the recovery of 70-75% of the MT originally added onto the filter.

2.2.4. Post-incubation of MT in the presence of dithiothreitol (DTT)

MT was incubated (under identical conditions to those described above) with either H,O,, H,O, /Fe2 + , Xan/XO, or AAPH, during 45, 45, 60 and 45 min, respectively. After stopping the incu- bation reactions, aliquots from each system were post-incubated during 60 min in the dark, at 20°C and in the presence of 2.5 mM DTT. Sam- ples were subsequently used to as& the “‘Cd- binding capacity. In the case of samples obtained from the AAPH experiments, the remaining azo compound was first removed as described above, and 30 ~1 of 2.5 mM DTT was subsequently added directly onto the filter. Before assaying for the Cd binding capacity, the protein was recov- ered following the same procedure. Appropriate controls were run in parallel.

2.3. Analytical procedures

2.3.1. Thiol content of MT Thiols were titrated using 5,5’-dithiobis(2-ni-

trobenzoic acid) (Nbs,) as follows: 100 ,~l of samples were removed from the incubation sys- tem and added into a cold Eppendorf tube. Sub- sequently, 100 ,U 1 of a 3 mM Nbs, solution were added and, after completing to a final volume of 1 ml with 100 mM potassium phosphate buffer, pH 7.4, the mixture was incubated during 60 min at 20°C. Optical density was read at 412 nm against a blank containing no MT. Thiolate con- centrations were calculated using an absorption coefficient of 13600 M-‘ecn-‘.

2.3.2. Cd-binding capacity Binding of radioactive Cd to MT was assayed

as described by Eaton and Toal (1982). Carrier- free “‘Cd was diluted with CdCl, . H,O prepared in 10 mM Tris buffer, pH 7.4, to yield a final Cd concentration of 2.0 pg/ml and an activity of 1 .O ,&i/ml. Two-hundred microliters of this solution were added to a tube containing 100 ,~l of the Tris buffer and 100 ~1 of the incubated sample (equivalent to 1 pg of MT). The mixture was incubated at 20°C during 10 min. To remove unbound “‘Cd, 100 ~1 of a 2% w/v freshly pre- pared bovine hemoglobin solution were added. The resulting mixture, contained in Eppendorf tubes whose caps had been punctured with a needle, was heated at 100°C during 2 min. After cooling, a second 100 ,~l aliquot of the hemoglobin solution was added, the mixture was heated as above, and subsequently centrifuged at 10 000 x g during 10 min. From the clear super- natant obtained, a 500 ,~l aliquot was removed, transferred to a gamma counting tube, and the radioactivity was measured. Blank tubes contain- ing buffer instead of MT were always run to determine the background counts. The total counts added to the system were determined in separate tubes containing “‘Cd but not MT or hemoglobin. The amount of MT (assayed as cad- mium binding capacity) in each sample was cal- culated according to the following equation:

nmol Cd = (Cts,, - Cts,,,) x 3.56/Cts,,

Cts (s): counts per minute in the sample tube; Cts (BKGJ: counts per minute in the background

tube; Cts (Tj: counts per minute in the ‘total count’

tube. The value of 3.56 was obtained by dividing the

grams of cadmium in each tube (0.4 fig) by the atomic weight of cadmium (112.4 g/mol)

2.4. Statistical analysis

Groups of test data (means + S.D.) were com- pared using Student’s t-test for paired observa- tions. Values were considered to differ significantly at P c-0.05.

42 I. Jim.knez et al. /Toxicology 120 (1997) 37-46

3. Results

3.1. Exposure of metallothionein to H,O, plus Fe=+

The incubation of MT with a mixture of Fe’+ (10 ,uM) and H,O, (500 ,uM) resulted in the progressive loss of its “‘Cd-binding capacity (Fig. 1A). After 45 min of exposure to this hy- droxyl radical generator, less than 26% of the initial binding capacity of the protein was recov- ered. This change was accompanied by a con- comitant and parallel loss of the thiol residues of MT (Fig. 1B). Both changes were closely and significantly (P = 0.001) correlated with an r

value of 0.998 (Fig. 1C).

3.2. Exposure of metallothionein to H,O, alone

The incubation of MT with H,O, (500 PM), in the absence of iron, resulted in losses of the lo9Cd binding capacity and the thiol residues of the protein identical (and therefore not shown here) to those observed when MT was incubated in the presence of the metal (corresponding to Fig. 1A and Fig. 1B). The changes observed in the absence of iron were closely and significantly (P = 0.001) correlated with an r value of 0.999 (not shown).

3.3. Exposure of metallothionein to

xanthinelxanthine oxidase

The continuous incubation of MT in presence of the superoxide radical generator xanthine/ xanthine oxidase led to a time-dependent de- crease of the “‘Cd-binding capacity. After 60 min of incubation the remaining capacity of MT to bind cadmium was about 19% (Fig. 2A). On the other hand, the thiolate content of metal- lothionein was observed to decay concomitantly (Fig. 2B). A direct and close relationship be- tween both parameters was established with a correlation coefficient of 0.993 (P = 0.001) (Fig. 2C).

80-

68-

40-

20-

o- 0 10 20 30 40 so 60 70

TIME (min)

looI

-20’ 0 20 40 60 80 loo

% DECREASE OF TEIOL!S

Fig. 2. Time course of the effects of the exposure of metalloth-

ionein to a xanthineixanthine oxidase system. Data are ex-

pressed as % of decrease of the “‘Cd-binding capacity (Fig.

2A), and % of decrease of thiol content (Fig. 2B) resulting

from incubating MT (0.1 mg/ml) in the presence of xanthine

oxidase (0.01 U/ml), in 100 mM potassium phosphate buffer,

pH 7.4. The reaction was initiated by the addition of xanthine to a final concentration of 1.0 mM. Each plotted point corre-

sponds to the mean of at least three independent experiments

(S.E.M. were always less than 5%). Assays were conducted as

described in Section 2. Fig. 2C depicts the correlation between

the % of decrease of the lWCd-binding capacity and its corre-

sponding % of decrease of thiol MT content. The line drawn is

based on a linear regression analysis (rZ = 0.986).

I. Jimhez et al. /Toxicology 120 (1997) 37-46 43

3.4. Effects of the addition of inhibitors on MT incubated in the presence of hydrogen peroxide or free radical-genenator systems

Table 1 represents the percentage of remaining metal binding calpacity and thiol content of MT after its brief incubation with either H202 (250 ,uM) alone (system A), H,O, (500 ,uM) plus Fe2 + (10 PM) (system B) or with Xan/XO (system C). The loss of cadmium binding capacity and thiol content induced by H,O, alone (system A) was not affected by the addition of SOD nor of desfer- rioxamine (DESF). In system B, the addition of catalase (CAT) led to complete prevention of the Fenton reaction-mediated loss of cadmium bind- ing capacity and %thiol content of MT. In turn, the addition of this enzyme to the Xan/XO reaction (system C) afforded a minor but not significant (P > 0.05) protection. On the other hand, while the addition of lSOD to the latter system com- pletely prevented MT from losing its metal bind- ing capacity and thiol content, the same enzyme afforded no protection against the effects induced by the Fenton reaction (system B). The addition of DESF had no effect on system B.

3.5. Exposure of metallothionein to AAPH

Incubation of MT in the presence of AAPH led to a progressive loss of the “‘Cd-binding capacity of MT. After 45 min of exposure to this peroxyl radical generator, less than 7% of the initial bind- ing capacity was recovered (Fig. 3A). This change was accompanied by a parallel loss of the thiol residues of MT (Fig. 3B). Both changes were closely and significantly (P = 0.002) correlated with an r value of 0.986 (Fig. 3C).

3.6. Effects of the post-incubation of oxidatively modl$ed MT with DTT

MT samples previously exposed to each of the above-described free radical generating systems were post-incubated in the presence of dithiothrei- to1 (as described in Section 2). The treatment of the oxidatively modified MT samples with DTT lead, in each case, to a full recovery (100%) of the capacity of MT to bind ‘@Cd, as seen before its

0 20 40 60 80 I

Yo DECREASE OF TEIOLS lo

Fig. 3. Time course of the effects of the exposure of metalloth- ionein to 2,2’-azobis(2-amidinopropane)dihydrochloride. Data are expressed as % of decrease of the ‘oQCd-binding capacity (Fig. 3A), and % of decrease of thiol content (Fig. 3B) resulting from incubating MT (0.1 mg/ml) in the presence of AAPH (25 mM), in 100 mM potassium phosphate buffer, pH 7.4. Each plotted point corresponds to the mean of at least three independent experiments (S.E.M. were always less than 5%). Assays were conducted as described in Section 2. Fig. 3C depicts the correlation between the % of decrease of the ‘09Cd-binding capacity and its corresponding % of decrease of thiol MT content. The line drawn is based on a linear regres- sion analysis (r’ = 0.972).

44 I. Jim&e: et al. /Toxicology 120 (1997) 37-46

incubation with any of the three free radical-gen- erating systems (data not shown).

4. Discussion

Recently, increasing attention has been given to metallothionein as a potential scavenger for chem- ically reactive species on the grounds of its partic- ularly high thiol content and its demonstrated role in cytoprotecting against physical conditions and chemical interventions associated with oxidative stress (Thornalley and Vasak, 1985; Thomas et al., 1986; Greenstock et al., 1987; Hidalgo et al., 1988; Mello-Filho et al., 1988; Satoh et al., 1988; Sato and Bremner, 1993). In the present study we addressed in vitro the interaction between metal- lothionein and the pro-oxidant species hydrogen peroxide, hydroxyl radical, superoxide anion, and peroxyl radical, in terms of their ability to oxidize the thiol residues of this protein. Such interaction was assessed with the additional purpose of estab- lishing a possible relationship between the antiox- idant or free radical scavenging properties of MT and its thiol-dependent capacity to bind and detoxify metals.

Our work demonstrates that the exposure of MT to hydrogen peroxide alone results in a con- centration- and time-dependent oxidation of its thiol residues. Since such an effect was not af- fected by the addition of an excess of SOD, superoxide anions, if present with the peroxide, would not seem to participate as mediators of the protein oxidation. On the other hand, the effect of the peroxide was not affected by the addition of catalytic amounts of iron, thus suggesting that hydroxyl radical species - expected to be gener- ated via the Fenton reaction - are not implicated in the peroxide-dependent oxidative modification of MT. Nonetheless, these results could also be taken as an indication that the hydroxyl radicals generated via the Fenton reaction are just as effective in oxidizing MT as the peroxide molecules from which they are derived. The latter interpretation, however, is contrary to our finding that the addition of high concentrations of manni- to1 (not shown), a well known hydroxyl radical scavenger, fails to prevent the effect of hydrogen

peroxide on MT, regardless of whether the perox- ide was added in the presence or absence of iron. While it could be argued that the oxidation of MT thiols could be the result of the reaction of minute amounts of hydroxyl radicals generated by catalytic concentrations of contaminating iron (despite chelex treatment of the solutions), the lack of protection by desferrioxamine strengthens the interpretation that the sole presence of perox- ide suffices to induce MT oxidation, and that - under the oxidizing conditions employed here - hydroxyl radicals do not participate or contribute to such an effect. We do not have an explanation for the lack of reactivity seen by us between hydroxyl radicals and MT thiols. Thus, our find- ings contrast with the original report by Thornal-

ley and Vasak, 1985 showing a higher reactivity of HO., relative to superoxide anions, towards MT. However, within the frame of our observations, the contention that hydroxyl radicals interact with MT is clearly not supported by our experimental data, at least, not to the extent that these species act inducing thiol oxidation of the protein, or the loss of the its metal-binding properties.

The interaction of MT with superoxide anions was evidenced using the xanthinelxanthine oxi- dase reaction in the presence of SOD. The addi- tion of SOD (in excess) to the Xan/XO system, to disproportionate superoxide anions into hydrogen peroxide, completely prevented thiol oxidation, supporting the view that superoxide anions are the only species implicated in the oxidative mod- ification of MT observed in this system. Since the addition of SOD is expected to substantially ele- vate the steady-state production of hydrogen per- oxide from superoxide anion, the complete inhibition of oxidation seen upon addition of this enzyme would appear to be in contradiction with the contention that MT oxidation is driven by the peroxide. However, a plausible explanation for such a result could be that, under our experimen- tal conditions, the actual concentration flux of superoxide anions generated via the Xan/XO reac- tion, while sufficient to induce its effects on MT, might be too low to allow accumulation of hydro- gen peroxide concentrations capable of promoting significant oxidation of the protein. On the other hand, the same result seems to indicate that super-

I. Jimhez et al. / Toxicology 120 (1997) 37-46 45

oxide anions - which in the Xan/XO reaction are generated equimolarly with hydrogen perox- ide, are appreciably more reactive towards MT thiols than the peroxide. The latter hypothesis was further confirmed by the complete lack of effect of catalase on the superoxide anion-driven MT oxi- dation. These findings are in line with the obser- vation by Fliss and MCnard (1992) showing that superoxide anions are extremely effective in pro- moting the mobilization of Zn atoms from MT, the latter presumably resulting from the oxidation of its thiol residues. Our findings are also in agreement with the early work by Hartmann et al. (1985) which implicated hydrogen peroxide as an oxidant of MT during its incubation in the pres- ence of neutrophiles. Hartmann et al. (1985), however, found superoxide anions not to be im- plicated in the oxidation of the protein. Finally, peroxyl radicals, which were generated directly as single species uict the thermal decomposition of AAPH, were also shown to induce the oxidative loss of MT thiol residues.

The observation that the loss of thiol groups in the MT molecule is accompanied by a concomi- tant and progressive loss of its lo9Cd-binding capacity, indicates that the interaction between MT and hydrogen peroxide, superoxide anions or peroxyl radicals leads not only to thiol-related structural changes in the protein, but also to the loss of one of its most relevant functional proper- ties. The demonstration that the capacity of MT to bind “‘%d, largely lost as a result of its prior exposure to the iabove mentioned species was, in all cases, completely recovered after post-incuba- tion with dithiotlhreitol, reveals the oxidative and reversible character of the peroxide and free radi- cal-mediated loss of thiols in the MT molecule. Furthermore, the full recovery of the native metal-binding pr’operties of MT suggests that the interaction between MT and hydrogen peroxide or free radicals does not involve either its frag- mentation or its non-oxidative polymerization. Thus, our findings are in line with the observation by K%gi (1991) in that all cysteines present in the MT molecule may occur in their reduced form and partake of .its metal binding properties. On the other hand, the strong and significant correla- tions found by us between the thiol content of

MT and its metal-binding capacity reveal that the latter property of MT is strictly dependent and directly proportional to the availability of thiol groups in the protein.

The oxidative modification of metallothionein via free radicals implicates the sulfhydryl residues of the protein - and its oxidability - as com- mon targets to mechanistically link the character- istic property of MT of binding metals with its more recently - under study - ability to scav- enge free radical species. Our study shows that both initiating (superoxide anions) as well as propagating (peroxyl radicals) oxygen-derived radical species, can directly interact with MT to oxidatively modify the protein. Although these results were obtained in vitro, they suggest that free radicals generated under in vivo conditions might also affect both properties of the protein. Such effects of free radicals on MT might become particularly relevant under oxidative conditions (such as nutrient and drug metabolism, inflamma- tory diseases, diabetes, hypoxia/reoxygenation, etc.). If significant oxidation of thiol residues of MT takes place in vivo, the cell’s capacity to handle those metals whose disposition depends on binding to MT is expected to be severely compro- mised. Furthermore, the latter condition could signify a particular risk of toxicity to those cells that are affected simultaneously by a stringent oxidative condition and an overload or exposure to such potentially toxic metals.

Acknowledgements

This work was supported by grants from COCHILCO (Chilean Copper Commision) and FONDECYT, project # 1950258.

References

Abel, J. and Ruiter, N. (1989) Inhibition of hydroxyl-radical generated DNA degradation by metallothionein. Toxicol. Lett. 47, 191-196.

Bakka, A., Johnsen, AS., Endresen, L. and Rugstad, H.E. (1982) Radioresistance in cells with high content of metal- lothionein. Experientia 38, 381-383.

46 I. Jimhez et al. / Toxicology 120 (1997) 37-46

Bray, T.M. and Bettger, W.J. (1990) The physiological role of

zinc as an antioxidant. Free Radic. Biol. Med. 8, 281-291.

Bremner, I. (1987) Interactions between metallothionein and

trace metals. Prog. Food Nutr. Sci. 11, l-37.

Cagen, X.Z. and Klaassen, C.D. (1979) Protection of carbon

tetrachloride (CC&)-induce hepatotoxicity by zinc: role of

metallothionein. Toxicol. Appl. Pharmacol. 51, 107-I 16.

Coppen, D.E., Richardson, D.E. and Cousins, R.J. (1988) Zinc

supression of free radicals induce in cultures of rats hepato-

cytes by iron, tert-butyl hydroperoxide, and 3-methylindole.

Proc. Sot. Exp. Biol. Med. 189, lOO- 109.

Cousins, R.J. (1985) Absorption, transport, and hepatic

metabolism of copper and zinc: special reference to metal-

lothionein and ceruloplasmin. Physiol. Rev. 65(2), 238-309.

Eaton, D.L. and Toal, B.F. (1982) Evaluation of the Cd/

hemoglobin affinity assay for the rapid determination of

metallothionein in biological tissues. Toxicol. Appl. Phar-

macol. 66, 134-142.

Fliss, H. and Menard, M. (1992) Oxidant-induced mobilization

of zinc from metallothionein. Arch. Biochem. Biophys.

293(l), 195-199.

Freedman, J.H. and Peisach, J. (1989) Resistence of cultured

hepatoma cells to copper toxicity. Purification and charac-

terization of the hepatoma metallothionein. Biochim. Bio-

phys. Acta 992, 1455154.

Goering, P.L. and Klassen, C.D. (1984) Zinc-induced tolerance

to cadmium hepatotoxicity. Toxicol. Appl. Pharmacol. 74,

2999307.

Greenstock, C.L., Jinot, C.P., Whitehouse, R.P. and Sargent,

M.D. (1987) DNA radiation damage and its modification by

metallothionein. Free. Radic. Res. Commun. 2, 233-239.

Hartmann, H.J., Gartner, A. and Wesser, U. (1985) Neu-

trophile-generated oxidants cleave metallothionein and in-

duce Zn/Cd/Cu release. Clin. Chim. Acta 152, 955103.

Hidalgo, J., Campany, L., Borras, M., Garvey, J.S. and Ar-

mario, A. (1988) Metallothionein response to stress in rats:

role in free radical scavenging. Am. J. Physiol. 255, E: 51882410.

Kagi, J. H. R. (1991) Overview of metallothionein. In: J.F.

Riordan and B.L. Vallee (Eds), Methods in Enzymology,

Vol. 205, Academic Press, San Diego, Ca., pp. 613-630.

Kagi, J. H. R, Hunziker, P. and Vaslk, M. (1990) Metalloth-

ionein: biochemistry and spatial structure. In: H. Tomita

(Ed), Trace Elements in Clinical Medicine, Springer-Verlag,

Tokyo, pp. 395-403.

Kennedy, M.C., Gan, T., Antholine, W.E. and Petering, D.H.

(1993) Metallothionein reacts with Fe*+ and NO to form

products with a g = 2.039 ESR signal. Biochem. Biophys.

Res. Comm. 196(2), 632-635.

Laze, J.S. and Pitt, B.R. (1995) Metallothioneins and cell death

by anticancer drugs. Annu. Rev. Pharmacol. Toxicol. 35,

6355653.

Lehman-McKeeman, L.D., Andrews, G.K. and Klassen, CD.

(1988) Mechanisms of regulation of rat hepatic metalloth-

ionein-I and metallothionein-II levels following administra-

tion of zinc. Toxicol. Appl. Pharmacol. 92, 1-9.

Leyshon-S$rland, K., M$rkrid, L. and Rugstad, H.E. (1993)

Metallothionein: a protein conferring resistance in vitro to

tumor necrosis factor. Cancer. Res. 53, 4874-4880.

Li, T.Y., Kraker, A.J., Shaw, C.F. III. and Petering, D.H.

(1980) Ligand substitution reactions of metallothioneins

with EDTA and apo-carbonic anhydrase. Proc. Natl. Acad.

Sci. USA 77, 633446338.

Mello-Filho, A.C., Chubatsu, L.S. and Meneghine, R. (1988)

V79 Chinese hamster cells rendered to high cadmium con-

centration also became resistant to oxidative stress.

Biochem. J. 256, 475-47934.

Renan, M.J. and Dowman, P.I. (1989) Increased resistance of

tumor cells exposed to metallothinein-inducing agents. Ra-

diat. Res. 120, 442-455.

Sato, M. and Bremner, I. (1993) Oxygen free radicals and

metallothionein. Free Radic. Biol. Med. 14, 325-337.

Satoh, M., Naganuma, A. and Imura, N. (1988) Involvement

of cardiac metallothionein in prevention of adriamicin

induce lipid peroxidation in the heart. Toxicology 53,

231-237.

Templeton, D.M. and Cherian, M.G. (1991) Toxicological

significance of metallothionein. In: J.F. Riordan and B.L.

Vallee (Eds), Methods in Enzymology, Vol. 205, Academic

Press, San Diego, Ca, pp. 11-35.

Thomas, J.P., Bachowski, G.L. and Girotti, A.W. (1986)

Inhibition of cell membrane lipid peroxidation by cadmium

and zinc-metallothionein. Biochim. Biophys. Acta 884,

448-461.

Thornalley, P.J. and Vaslk, M. (1985) Possible role for metal-

lothionein in protection against radiation-induce oxidative

stress. Kinetics and mechanism of its reaction with superox-

ide and hydroxyl radicals. Biochim. Biophys. Acta. 827,

36-44.