MP

MJLS

a

ARRA

KICLTCN

1

eiwwAgba

mPOT

0d

Chemico-Biological Interactions 188 (2010) 220–227

Contents lists available at ScienceDirect

Chemico-Biological Interactions

journa l homepage: www.e lsev ier .com/ locate /chembio int

echanisms underlying iron and copper ions toxicity in biological systems:ro-oxidant activity and protein-binding effects

aría Eugenia Letelier ∗, Sebastián Sánchez-Jofré, Liliana Peredo-Silva,uan Cortés-Troncoso, Paula Aracena-Parksaboratory of Pharmacology and Toxicology, Department of Pharmacological and Toxicological Chemistry,chool of Chemical and Pharmaceutical Sciences, Universidad de Chile, Santiago 8380492, Chile

r t i c l e i n f o

rticle history:eceived 4 February 2010eceived in revised form 11 May 2010ccepted 24 June 2010

eywords:ronopperipid peroxidationhiol proteinsatechin-Acetylcysteine

a b s t r a c t

Iron and copper ions, in their unbound form, may lead to the generation of reactive oxygen speciesvia Haber–Weiss and/or Fenton reactions. In addition, it has been shown that copper ions can irre-versibly and non-specifically bind to thiol groups in proteins. This non-specific binding property hasnot been fully addressed for iron ions. Thus, the present study compares both the pro-oxidant and thenon-specific binding properties of Fe3+ and Cu2+, using rat liver cytosol and microsomes as biological sys-tems. Our data show that, in the absence of proteins, Cu2+/ascorbate elicited more oxygen consumptionthan Fe3+/ascorbate under identical conditions. Presence of cytosolic and microsomal protein, however,differentially altered oxygen consumption patterns. In addition, Cu2+/ascorbate increased microsomallipid peroxidation and decreased cytosolic and microsomal content of thiol groups more efficiently thanFe3+/ascorbate. Finally, Fe3+/ascorbate and Cu2+/ascorbate inhibited in different ways cytosolic and micro-somal glutathione S-transferase (GST) activities, which are differentially sensitive to oxidants. Moreover,in the absence of ascorbate, only Cu2+ decreased the content of cytosolic and microsomal thiol groups

and inhibited cytosolic and microsomal GST activities. Catechin partially prevented the damage to thiolgroups elicited by Fe3+/ascorbate and Cu2+/ascorbate but not by Cu2+ alone. N-Acetylcysteine completelyprevented the damage elicited by Cu2+/ascorbate, Fe3+/ascorbate and Cu2+ alone. N-Acetylcysteine alsocompletely reversed the damage to thiol groups elicited by Fe3+/ascorbate, partially reversed that ofCu2+/ascorbate but failed to reverse the damage promoted by Cu2+ alone. Our data are discussed in termsto the potential damage that the accumulation of iron and copper ions can promote in biological systems.

. Introduction

In biological systems, iron and copper ions are essential forlectron-transfer reactions [1,2]. Dietary intake of these metalss, therefore, necessary for all living organisms [3–5]. However,

hen these ions are in their free unbound form, they can interactith oxygen by catalyzing Haber–Weiss and/or Fenton reactions.

s a consequence of such reactions, reactive oxygen species areenerated which, uncontrolled, can lead to oxidative damage toiomolecules [6,7]. All living systems have evolved several mech-nisms that allow transport and storage of iron and copper ions in

∗ Corresponding author at: Laboratory of Pharmacology and Toxicology, Depart-ent of Pharmacological and Toxicological Chemistry, School of Chemical and

harmaceutical Sciences, Universidad de Chile, Sergio Livingstone Pohlhammer (ex-livos) 1007, Independencia, Santiago 8380492, Región Metropolitana, Chile.el.: +56 2 9782885; fax: +56 2 9782996.

E-mail address: mel@ciq.uchile.cl (M.E. Letelier).

009-2797/$ – see front matter © 2010 Elsevier Ireland Ltd. All rights reserved.oi:10.1016/j.cbi.2010.06.013

© 2010 Elsevier Ireland Ltd. All rights reserved.

order to keep them in a bound form [1,2]. Physiological plasma ironand copper total concentrations (free and bound) range between0.6–1.7 mg/L (∼11–31 �M) and 0.9–1.2 mg/L (∼14–19 �M), respec-tively [8–10]. Overload of iron and/or copper ions can be due tohigh dietary intake or alterations in their transport. This conditionmay overwhelm the ability of cells to keep these ions protein-bound [4,11]. Several pathologies, that include neurodegenerativediseases, such as Parkinson’s and Alzheimer’s [12–15], and hepaticdisorders, such as haemochromatosis and Wilson’s disease [16,17]are associated to the overload of iron and/or copper ions. Therefore,it is widely accepted that the damage observed in such patholo-gies is the consequence of the pro-oxidant properties of these ions[17,18].

We have reported that copper ions can bind to proteins non-

specifically [19], leading to the alteration of several liver enzymesinvolved in xenobiotic biotransformation (namely cytochromeP450 oxidative system, UDP-glucuronyltransferase, and glu-tathione S-transferases) [19–22]. In these studies, �M Cu2+, but notnM Cu2+, was shown to elicit significant binding to proteins [19].

gical

Taoia

fppsptwtptcctiebaoatm

bt(sotaotohomab

2

2

OcD(SCwHi

2

tea

M.E. Letelier et al. / Chemico-Biolo

his binding property of copper ions may be shared with iron ions,lthough this hypothesis has yet to be tested. Relative contributionf pro-oxidant activities and non-specific binding of iron or copperons to biomolecules may be relevant in diseases associated to theirccumulation.

In this work, we postulated that Fe3+ and Cu2+ will elicit dif-erential damage to biological systems due to their particularro-oxidant activities and non-specific binding properties. In theresence of ascorbate, these ions will generate reactive oxygenpecies [23], in addition to their potential non-specific binding toroteins [19]. On the other hand, in the absence of ascorbate, solelyhe non-specific binding effect should be observed [19]. Therefore,e first compared the ability of Fe3+/ascorbate and Cu2+/ascorbate

o consume molecular oxygen, which was also evaluated in theresence of rat liver cytosol or microsomes. To evaluate damageo biomolecules, we also used liver cytosol and microsomes, whichontain the main biotransformation systems. These systems areonstituted by thiol enzymes, a structural condition that rendershem sensitive to their redox environment and potential bind-ng of iron and/or copper ions. Thus, we compared the damagelicited by Fe3+ and Cu2+, in the presence or absence of ascorbate,y assessing microsomal lipid peroxidation and the loss of cytosolicnd microsomal thiol contents. Catechin and N-acetylcysteine arexygen-free radical scavengers and/or iron or copper ions chelatinggents. Thus, we evaluated if these agents could prevent or reversehe alterations promoted by iron and copper ions on cytosolic and

icrosomal thiol contents.To address if the oxidative and/or binding phenomena promoted

y Fe3+ and Cu2+ are correlated with a change in biomolecules func-ion, we also assessed the activities of glutathione S-transferaseGST). These isoenzymes, which occur in hepatic cytosol and micro-omes, are differentially sensitive to oxidants [22,24]. Promotionf disulfide-bound dimers, through direct oxidation of the pro-ein, decreases cytosolic GST activity and increases microsomal GSTctivity [24–26]. Microsomal GST activity, however, is decreased ifxidation occurs in its lipid environment rather than directly onhe protein [24]. We have shown that copper ions, in the absencef ascorbate, decrease microsomal GST activity, presumably byindering the formation of catalytically active dimers [19]. Databtained are discussed in terms to the potential damage that accu-ulation of iron and copper ions can promote in biological systems

nd the protection that can be provided by polyphenol- and thiol-ased antioxidants.

. Materials and methods

.1. Chemicals

1-Chloro-2,4-dinitrobenzene was purchased from ACROSrganics (New Jersey, NJ, USA). Bovine serum albumin (BSA),atechin, 5,5′-dithiobis-(2-nitrobenzoic acid) (Ellman’s reagent,TNB), iminodiacetic acid sodium form in polystyrene matrix

CHELEX-100), and N-acetylcysteine (NAC) were obtained fromigma Chemical Co. (St. Louis, MO, USA). CuSO4, FeCl3, Folin-iocalteu’s reagent, sodium ascorbate, and thiobarbituric acid (TBA)ere obtained from Merck Chile. All solutions were prepared inPLC-water previously treated with CHELEX-100. All other chem-

cals used were of analytical grade.

.2. Animals

Adult male Sprague–Dawley rats (200–250 g), maintained athe vivarium of the School of Chemical and Pharmaceutical Sci-nces (Universidad de Chile, Santiago, Chile) were used. Rats werellowed free access to pelleted food, maintained with controlled

Interactions 188 (2010) 220–227 221

temperature (22 ◦C) and photoperiod (lights on from 07:00 to19:00 h). All procedures were performed using protocols approvedby the Institutional Ethical Committee of the School of Chemicaland Pharmaceutical Sciences, Universidad de Chile, and accordingto the guidelines of the Guide for the Care and Use of LaboratoryAnimals (NRC, USA).

2.3. Isolation of liver cytosol and microsomes

Cytosolic and microsomal fractions were prepared according toLetelier et al. [19,22]. Total protein was determined according toLowry et al. [27].

2.4. Determination of oxygen consumption

Oxygen consumption was continuously determined polaro-graphically up to 2 min with a Clark Electrode No. 5300 (YellowsSprings Instruments Co., Inc., Yellow Springs, OH, USA) coupledwith a Windaq transducer model DATAQ D1-148U. Oxygen con-centrations were calculated considering 100% oxygen saturationas 452.5 nmol/mL (at the local atmospheric pressure, which corre-sponds to 718 mm Hg).

2.5. Assay of microsomal lipid peroxidation

The extent of microsomal lipid peroxidation was estimated bydetermining TBARS, according to Letelier et al. [19].

2.6. Determination of thiol content

Thiol groups were titrated with DTNB as described by Letelieret al. [19,22].

2.7. Assay of GST activity

GST activities were assayed using 1-chloro-2,4-dinitrobenzeneand GSH as substrates, as described by Letelier et al. [19,22]. Non-linear regression of the data was performed with the equationA = C + AmaxIC50/([ion] + IC50), where A corresponds to the experi-mental GST activity at the ion concentration [ion], Amax correspondsto the %inhibition of GST activity at the highest [ion] tested, IC50 cor-responds to the half-maximum inhibition of the GST activity, and Ccorresponds to the residual GST activity at the highest [ion] tested.

2.8. Statistical analysis

Data presented in this work correspond to the arithmeticalmean of at least 4 independent experiments ± SEM. Statistical sig-nificance (ANOVA, followed by post hoc Bonferroni) and regressionanalyses were performed using GraphPad Prism 5.0. Values wereconsidered to differ significantly at the level of p < 0.05.

3. Results

3.1. Oxygen consumption

Fig. 1A depicts representative traces of oxygen consumptionelicited by either Fe3+ or Cu2+ in the absence of ascorbate (trace a),or following the addition of 1 mM ascorbate (traces b and c). We cal-culated the rates of oxygen consumption from linear regression ofthe data. Notably, 50 �M Cu2+/ascorbate promoted approximately

50-fold more oxygen consumption than 50 �M Fe3+/ascorbate infree solution (Fig. 1B and C). These oxygen consumption patternschanged differentially when cytosolic or microsomal preparations(1 mg protein/mL) were added to the mixture. Microsomes (butnot cytosol) increased oxygen consumption by Fe3+/ascorbate in

222 M.E. Letelier et al. / Chemico-Biological Interactions 188 (2010) 220–227

Fig. 1. Oxygen consumption elicited by Cu2+/ascorbate and Fe3+/ascorbate: effect ofcytosolic and microsomal protein. Oxygen consumption elicited by Cu2+/ascorbateor Fe3+/ascorbate was measured as detailed in Section 2. Representative tracesof oxygen consumption in the presence of 50 �M Fe3+ or Cu2+ in the absence ofascorbate (Control, trace a), and following addition of 1 mM ascorbate (traces band c) are plotted in A. Rates of oxygen consumption by either Fe3+/ascorbate (B)or Cu2+/ascorbate (C) were obtained from the slopes of the curves following theaddition of 1 mM ascorbate, in the absence or presence of 1 mg/mL cytosolic ormdc

admFt

3m

nhpml(p

Table 1Microsomal lipid peroxidation elicited by Cu2+ and Fe3+.

System TBARS, nmol/min/mg proteinNone 0.03 ± 0.001

With 1 mM ascorbate Without 1 mM ascorbate

25 nM Fe3+ 0.03 ± 0.001 ND50 �M Fe3+ 0.19 ± 0.033a ND25 nM Cu2+ 0.12 ± 0.003a ND50 �M Cu2+ 0.04 ± 0.001b ND

Lipid peroxidation was estimated by measuring TBARS following incubation ofmicrosomes (1 mg/mL) with Fe3+ or Cu2+, in the presence or absence of 1 mMascorbate, for 30 min at 37 ◦C. Basal microsomal lipid peroxidation was measured

3+ 2+

detectable changes in cytosolic or microsomal GST activities evenat the highest concentration tested (400 �M). In contrast, Cu2+

promoted a concentration-dependent inhibition of both GST activ-ities (Fig. 3). The highest Cu2+ concentration employed (400 �M)

Table 2Loss of thiol content elicited by Cu2+ and Fe3+.

System Sample Residual thiol content, %

With 1 mM ascorbate Without 1 mMascorbate

25 nM Fe3+ Cytosol 99.9 ± 0.3 99.3 ± 0.7Microsomes 101.2 ± 0.5 99.9 ± 0.7

50 �M Fe3+ Cytosol 46.5 ± 0.4a 101.3 ± 0.6Microsomes 68.7 ± 1.3b 99.8 ± 0.5

25 nM Cu2+ Cytosol 100.7 ± 1.1 99.4 ± 0.6Microsomes 100.1 ± 0.9 101.1 ± 0.4

50 �M Cu2+ Cytosol 23.9 ± 1.1b 37.6 ± 0.6b

Microsomes 2.2 ± 0.9b 23.4 ± 0.6b

Thiol content was titrated with DTNB following incubation of microsomes (1 mg/mL)with Fe3+ or Cu2+, in the presence or absence of 1 mM ascorbate, for 30 min at 37 ◦C.

icrosomal protein. Rate values correspond to the mean ± SEM of at least 4 indepen-ent determinations and are expressed as nmol O2 consumed/mL/min. ***p < 0.001ompared to control with no added protein.

bout 5.8-fold (Fig. 1B). In contrast, both cytosol and microsomesecreased oxygen consumption by Cu2+/ascorbate to approxi-ately the same extent (37% and 44%, respectively; Fig. 1C). Neither

e3+ nor Cu2+ alone (up to 400 �M) promoted any oxygen consump-ion (data not shown).

.2. Microsomal lipid peroxidation and loss of cytosolic andicrosomal thiol content

As shown in Table 1, basal microsomal lipid peroxidation wasot significantly altered by 25 nM Fe3+/ascorbate. On the otherand, 50 �M Fe3+/ascorbate promoted a ∼6.3-fold increase in thisarameter. In contrast, 25 nM Cu2+/ascorbate increased microso-

al lipid peroxidation in about 4-fold while 50 �M Cu2+/ascorbate

ed to a slight, but statistically significant, increase in this parameterabout 1.3-fold, Table 1). Neither Fe3+ nor Cu2+ alone (up to 400 �M)romoted any microsomal lipid peroxidation (data not shown).

in samples incubated without Fe , Cu or ascorbate. Values correspond to themean ± SEM of at least 4 independent determinations. ND: not detected.

a p < 0.05 compared to basal microsomal lipid peroxidation.b p < 0.001 compared to basal microsomal lipid peroxidation.

Incubating cytosolic and microsomal samples with 50 �MFe3+/ascorbate elicited approximately 55% and 30% loss of thiolcontent, respectively (Table 2). Under the same conditions, 50 �MCu2+/ascorbate promoted 76% and 98% loss of cytosolic and micro-somal thiol content, respectively (Table 2). Incubation of cytosolicand microsomal samples with 50 �M Cu2+ elicited 62% and 76% lossof thiol content, respectively. Incubation of samples with 50 �MFe3+ under the same conditions, however, promoted a negligibleeffect on this parameter (Table 2).

3.3. Inhibition of GST activities

Fe3+ and Cu2+, in the presence of ascorbate, led to the inhi-bition of cytosolic GST activity in a manner dependent with theion concentration (Fig. 2A). The highest concentration employedof Fe3+ and Cu2+ (400 �M) led to ∼30% and ∼80% inhibition ofthis activity, respectively. Fe3+/ascorbate displayed an IC50 valueapproximately 150-fold higher than Cu2+/ascorbate (Table 3). Like-wise, microsomal GST activity was decreased by Cu2+/ascorbateand Fe3+/ascorbate. However, this decrease was dependent on Cu2+

but not Fe3+ concentrations (Fig. 2B). In this case, the highest Fe3+

and Cu2+ concentrations used (400 �M) led to approximately 60%inhibition of microsomal GST activity in both cases (Fig. 2B).

As shown in Fig. 3A and B, Fe3+ alone failed to elicit any

Values correspond to the mean ± SEM of at least 4 independent determinations.Initial thiol content (100%) was measured in samples incubated without Fe3+, Cu2+

or ascorbate.a p < 0.01 compared to basal microsomal lipid peroxidation.b p < 0.001 compared to basal microsomal lipid peroxidation.

M.E. Letelier et al. / Chemico-Biological Interactions 188 (2010) 220–227 223

Fig. 2. Effect of Fe3+/ascorbate and Cu2+/ascorbate on cytosolic and microsomal GST actdifferent concentrations (up to 400 �M) of either Fe3+ (open circles) or Cu2+ (solid circles),(B) GST activities were assayed as described in Section 2. Values correspond to the mean ±activity; total GST activity was measured in samples incubated without Fe3+, Cu2+ or asco

Table 3Inhibition of GST activities by Fe3+ and Cu2+.

System Sample IC50 For GST inhibition

With 1 mMascorbate

Without 1 mMascorbate

50 �M Fe3+ Cytosol 115 (17.1–213.7) NDMicrosomes NC ND

50 �M Cu2+ Cytosol 0.76 (0.51–1.00)a 11.8 (5.1–17.9)Microsomes 118 (45.5–190.6) 58.0 (21.8–94.2)

GST activities were measured following incubation of cytosolic and microsomalsamples (1 mg/mL) with different concentrations (up to 400 �M) of either Fe3+

or Cu2+, in the presence of 1 mM ascorbate, for 30 min at 37 ◦C. IC50 values wereobtained from non-linear regression of the data, as detailed in Section 2, and corre-spond to the mean of at least 4 independent determinations, with the 95% confidenceintervals in parenthesis. NC: Not calculated, ND: not detectable.

lGImnn

incubation with NAC, however, completely reversed the loss ofcytosolic and microsomal thiol content elicited by Fe3+/ascorbate

Fw(a

a p < 0.001 compared to Fe3+/ascorbate.

ed to ∼35% and ∼60% inhibition of cytosolic and microsomalST activities, respectively. Cu2+ inhibited cytosolic GST with an

C50 value approximately 5-fold lower than that for inhibiting3+

icrosomal GST activity (Table 3). Since Fe was ineffective to sig-

ificantly inhibit either GST activity, these values were denoted asot detectable (Table 3, ND).

ig. 3. Effect of Fe3+ and Cu2+, in the absence of ascorbate, on cytosolic and microsomal Gith different concentrations (up to 400 �M) of either Fe3+ (open circles) or Cu2+ (solid ci

B) GST activities were assayed as described in Section 2. Values correspond to the mean ±ctivity; total GST activity was measured in samples incubated without Fe3+ or Cu2+.

ivities. Following incubation of cytosolic and microsomal samples (1 mg/mL) within the presence of 1 mM ascorbate, for 30 min at 37 ◦C, cytosolic (A) and microsomalSEM of at least 4 independent determinations and are expressed as % residual GST

rbate.

3.4. Effect of catechin and N-acetylcysteine (NAC) on the loss ofthiol groups elicited by Fe3+ and Cu2+ in the presence and absenceof ascorbate

Pre-incubation of cytosol and microsomes with catechin com-pletely prevented the loss of cytosolic thiol content and partiallyprevented the loss of microsomal thiol content elicited byFe3+/ascorbate (Fig. 4A, p < 0.05). On the other hand, catechin pre-incubation partially prevented the cytosolic and microsomal thiolcontent loss promoted by Cu2+/ascorbate to a much lower extentthan the prevention observed with Fe3+/ascorbate (Fig. 4A). In theabsence of ascorbate, only Cu2+ significantly decreased cytosolicand microsomal thiol contents (Table 2). Pre-incubation with cate-chin did not prevent this effect (Fig. 4B). In contrast, pre-incubationwith NAC completely prevented the loss of thiol content elicited byFe3+/ascorbate, Cu2+/ascorbate or Cu2+ alone (Fig. 4C and D, p > 0.05compared to control).

Fig. 5 shows that post-incubation with catechin failed to reversethe loss of microsomal and cytosolic thiol content elicited byFe3+/ascorbate, Cu2+/ascorbate or Cu2+ alone (Fig. 5A and B). Post-

(Fig. 5C). Moreover, NAC partially reversed the loss of cytosolic andmicrosomal thiol content promoted by Cu2+/ascorbate (Fig. 5C) andfailed to reverse that of Cu2+ alone (Fig. 5D).

ST activities. Following incubation of cytosolic and microsomal samples (1 mg/mL)rcles), in the absence of ascorbate, for 30 min at 37 ◦C, cytosolic (A) and microsomal

SEM of at least 4 independent determinations and are expressed as % residual GST

224 M.E. Letelier et al. / Chemico-Biological Interactions 188 (2010) 220–227

Fig. 4. Loss of cytosolic and microsomal thiol content elicited by Fe3+/ascorbate and Cu2+/ascorbate: prevention by catechin and NAC. Cytosolic and microsomal sampleswere pre-incubated with 0.1 mM of either catechin (A and B) or NAC (C and D) for 15 min at room temperature prior to incubation with 50 �M of either Fe3+ or Cu2+, in thepresence (A and C) or absence (B and D) of 1 mM ascorbate, for 30 min at 37 ◦C. Thiol groups were then titrated as detailed in Section 2. Values correspond to the mean ± SEMo t; tot 3+ 2+

a hiol coi ith catc

4

irtdApmatviinTn

f at least 4 independent determinations and are expressed as % residual thiol contennd is indicated by a dotted line. Hatched portions of the bars indicate the residual tllustrate the % of protection of residual thiol content from samples pre-incubated wompared to residual thiol content without pre-incubation with catechin or NAC.

. Discussion

It is widely acknowledged that the toxicity of iron and copperons is associated to their pro-oxidant properties. Studies of theedox effects elicited by these ions on liver biotransformation sys-ems have shown that Cu2+ concentrations promoting oxidativeamage are significantly lower than those of Fe3+ [19–22,24,28].lthough both are transition metals, they display different redoxotentials as well as coordination chemistries. Such differencesay lead to different toxicity mechanisms that underlie the dam-

ge promoted by these ions. In this regard, copper ions, in additiono their pro-oxidant activity [29], can non-specifically and irre-ersibly bind to thiol proteins altering their activity [19–22]. There

s a lot of information regarding the pro-oxidant activity of ironons [11,30,31]. However, the potential ability of such ions toon-specifically bind thiol proteins has not been fully addressed.herefore, in the present work, we evaluated the pro-oxidant andon-specific binding properties of Fe3+ and Cu2+. To this end, we

al thiol content was measured in samples incubated without Fe , Cu or ascorbatentent of samples without pre-incubation with catechin or NAC while grey portionsechin or NAC. **p < 0.05 and ***p < 0.001 compared to total thiol content; †††p < 0.001

tested the effect of Fe3+ and Cu2+, in the presence or absence ofascorbate, on rat liver microsomal lipids (substrates for oxidation)and protein thiol groups (substrates for oxidation and ligands ofmetal ions) occurring in rat liver cytosol and microsomes. Fur-thermore, we also assayed the effect of Fe3+ and Cu2+ on cytosolicand microsomal GST isoenzymes, which are thiol proteins differ-entially sensitive to oxidative damage and thiol group modification[19,22,24].

Studies of the oxygen consumption elicited by Fe3+ and Cu2+ inthe presence of ascorbate allow the evaluation of the generation ofreactive oxygen species (ROS), the true agents promoting biologi-cal oxidative damage. Presence of ROS biological substrates shouldincrease the rate of oxygen consumption due to the consumption

of ROS (Fig. 1). This seems to be the case of the increase in the rateof oxygen consumption by Fe3+/ascorbate elicited by microsomes:ROS generated by this system are consumed in microsomal lipidperoxidation. By the same token, in our assay conditions cytosolicproteins do not appear to contain good substrates for ROS gen-

M.E. Letelier et al. / Chemico-Biological Interactions 188 (2010) 220–227 225

Fig. 5. Loss of cytosolic and microsomal thiol content elicited by Fe3+/ascorbate and Cu2+/ascorbate: reversion by catechin and NAC. Cytosolic and microsomal samples wereincubated with 50 �M of either Fe3+ or Cu2+, in the presence (A and C) or absence (B and D) of 1 mM ascorbate, for 30 min at 37 ◦C. Samples were then incubated with 0.1 mMof either catechin (A and B) or NAC (C and D) for 15 min at room temperature prior to titrating thiol groups as described in Section 2. Values correspond to the mean ± SEM ofat least 4 independent determinations and are expressed as % residual thiol content; total thiol content was measured in samples incubated without Fe3+, Cu2+ or ascorbatea al thiop cubat† chin o

edaiodsbpr[

m(ctge

nd is indicated by a dotted line. Hatched portions of the bars indicate the residuortions illustrate the % of recovery of residual thiol content from samples further in††p < 0.001 compared to residual thiol content without further incubation with cate

rated by 50 �M Fe3+/ascorbate for 2 min. This difference may beue to the occurrence of lipids in microsomal preparations, whichppear to be better substrates for oxidative damage and are absentn cytosol. On the other hand, if a biological molecule can bind Fe3+

r Cu2+ ions, oxygen consumption should be decreased due to theecrease in their effective concentration. Both cytosol and micro-omes promoted a decrease in the rate of oxygen consumptiony Cu2+/ascorbate, suggesting that biomolecules occurring in bothreparations can bind to these ions. This postulate is supported byadioisotopic Cu2+ binding studies already reported by our group19].

Both Fe3+ and Cu2+ (in the presence of ascorbate) promotedicrosomal lipid peroxidation, but at different concentrations

2+ 3+

Table 1). While nM Cu elicits significant lipid peroxidation, a Feoncentration three orders of magnitude higher (�M) was requiredo observe this phenomenon. This is in agreement with the oxy-en consumption data, which shows that Cu2+/ascorbate is a morefficient ROS-generating system in the absence of cytosol or micro-

l content of samples without further incubation with catechin or NAC while greyed with catechin or NAC. **p < 0.05 and ***p < 0.001 compared to total thiol content;r NAC.

somes (Fig. 1). Remarkably, high (�M) concentrations of Cu2+ led tolower microsomal lipid peroxidation than low (nM) concentrations.This apparent paradox may be explained in terms of the effectiveCu2+ concentrations in each case. At �M Cu2+ concentrations, sig-nificant binding of this ion to biomolecules occur, decreasing Cu2+

effective concentration. At this concentration lipid peroxidationstill occurs (about 1.3-fold increase over basal lipid peroxidation),showing that both oxidative and binding phenomena are takingplace. This is also supported by the oxygen consumption data andby previously reported results using radiolabeled copper [19].

Loss of cytosolic and microsomal thiol content elicited by Fe3+

and Cu2+ is another example showing their differential nature interms of redox potential and binding properties. In the presence of

ascorbate (Table 2), both ions decreased cytosolic and microsomalthiol content but Cu2+ was more efficient to elicit this phenomenon.Remarkably, only Cu2+ decreased cytosolic and microsomal thiolcontent in the absence of ascorbate (Table 2). This suggests that,at least up to �M concentrations, only Cu2+ – but not Fe3+ – can

2 gical

ba

dcf[iainaainaCsmdllFaosiheFpib(ihaiti

iapprsdcmpaRcCrtrsaPCiiHp

[

[

[

[

[

[

26 M.E. Letelier et al. / Chemico-Biolo

ind non-specifically to amino acid residues occurring in proteins,postulate that is consistent with previous studies [19–22].

GST isoenzymes are fully active in a dimeric form bridged by aisulfide bond. Monomeric cytosolic GST isoenzymes display 3–4ysteine residues in their structure, whose oxidation leads to theormation of several inactive dimers and one possible active dimer22,24]. Therefore, non-specific oxidation elicited by Fe3+ and Cu2+

n the presence of ascorbate leads to the inhibition of cytosolic GSTctivity [22,24]. In contrast, the most abundant monomeric GSTsoenzyme (mGST1) display only one cysteine residue and, thus,on-specific oxidation leads to the activation of microsomal GSTctivity [19,24]. We have shown, however, that microsomal GSTctivity is strongly inhibited by lipid peroxidation, to the point thatts oxidative activation can only take place in the absence this phe-omenon [24]. This evidence suggests that changes in cytosolicctivity should be consistent with the effects produced by Fe3+ andu2+ on the thiol groups occurring in cytosol. Changes in micro-omal GST activity, however, should relate to both changes inicrosomal lipid peroxidation and microsomal thiol content. Our

ata show that, in the absence of ascorbate, the loss in cytoso-ic and microsomal GST activities completely correlated with theoss of thiol content elicited by copper ions (Table 2 and Fig. 3).urthermore, both activities were not altered by iron ions in thebsence of ascorbate, which also correlated with its negligible effectn thiol content (Table 2 and Fig. 3). Lipid peroxidation, however,eemed to display a more profound effect on microsomal GST activ-ty since Fe3+/ascorbate elicited a decrease in this activity to aigher extent than Cu2+/ascorbate (Table 1 and Fig. 3). All thesevidences support the following interpretations of our data: (1)e3+/ascorbate inhibits cytosolic GST activity through an oxidativerocess; (2) since Cu2+/ascorbate can inhibit cytosolic GST activ-

ty by direct oxidation of the protein plus non-specific binding, itehaves as a stronger inhibitor of this activity than Fe3+/ascorbate;3) Cu2+/ascorbate can promote inhibition of microsomal GST activ-ty through non-specific binding to microsomal thiol groups, whichinders the formation of the catalytically active dimer and alsolters the microsomal membrane conformation; (4) Fe3+/ascorbatenhibits microsomal GST activity altering its lipid environmenthrough microsomal lipid peroxidation, which makes it a strongernhibitor of microsomal GST activity than Cu2+/ascorbate.

Insofar, our data strongly suggest that mechanisms by whichron and copper ions are strongly related to their pro-oxidantctivities and non-specific binding properties. Catechin, a herbalolyphenol compound, and N-acetylcysteine (NAC), a thiol com-ound, are chelating agents of metal ions and are also oxygen-freeadical scavengers. Both agents prevented cytosolic and micro-omal thiol loss elicited by Fe3+/ascorbate and Cu2+/ascorbate toifferent extents (Fig. 4). Moreover, only NAC (but not catechin)ompletely prevented the cytosolic and microsomal thiol loss pro-oted by Cu2+ alone (Fig. 4). This suggests that NAC may be

rotecting protein thiol groups by acting as a chelating agent andROS scavenger, while catechin is likely to be acting mainly as aOS scavenger. Furthermore, catechin failed to reverse the loss ofytosolic and microsomal thiol content by either Fe3+/ascorbate,u2+/ascorbate, or Cu2+ alone (Fig. 5). NAC, however, completelyeversed the loss of thiol content elicited by Fe3+/ascorbate, par-ially reversed that promoted by Cu2+/ascorbate, but failed toeverse that promoted by Cu2+ alone (Fig. 5). Reversion datatrongly suggest that NAC may be acting as a thiol group reducinggent, a property that it does not appear to share with catechin.artial reversion of NAC on the loss of thiol content elicited by

u2+/ascorbate may be a reflection of the irreversible copper bind-

ng event. In summary, the non-specific binding property of copperons adds to their pro-oxidant activity in damaging biomolecules.owever, iron ions seem to only promote damage through theirro-oxidant activity, most likely due to a lower affinity towards

[

[

Interactions 188 (2010) 220–227

ligands occurring in proteins. Thus, it is possible to postulate thatin biological systems, copper overload may lead to more extensivetoxic consequences than iron overload. This postulate is currentlyunder evaluation.

5. Conclusion

Our data show that, under identical in vitro assay conditions,Cu2+ display more extensive toxic effects than Fe3+, in terms ofoxidative damage and non-specific binding to biomolecules. Fur-thermore, this difference is mainly due to the ability of copper ionsto non-specifically bind to protein thiol groups in addition to itspro-oxidant activity. These properties should be taken into accountwhen evaluating the toxicity of these ions and potential tools toprevent or reverse the damage they promote. The use of antioxi-dants to supplement the treatment of patients suffering differentpathologies that are associated to iron and/or copper overload andoxidative stress [12–17], such as Parkinson’s and Alzheimer’s dis-eases, is widely accepted [32,33]. Although plasma concentrationsof iron and copper ions reported for this type of diseases are con-sistent with the concentrations used in our study, it is necessaryto confirm the correlations found with in vivo studies. Our datahighlight the need of antioxidant drugs that not only can scavengeROS but also can chelate ions such as copper and iron [34,35]. Thiol-based antioxidants rather than polyphenol-based antioxidants mayfulfill both requirements; this is the subject of our continuingresearch.

Conflict of interest

The authors declare that there are no conflicts of interest.

References

[1] G.J. Anderson, C.D. Vulpe, Mammalian iron transport, Cell. Mol. Life Sci. 66 (20)(2009) 3241–3261.

[2] J.H. Kaplan, S. Lutsenko, Copper transport in mammalian cells: special care fora metal with special needs, J. Biol. Chem. 284 (38) (2009) 25461–25465.

[3] M. Arredondo, M.T. Nunez, Iron and copper metabolism, Mol. Aspects Med. 26(4–5) (2005) 313–327.

[4] M.C. Linder, M. Hazegh-Azam, Copper biochemistry and molecular biology, Am.J. Clin. Nutr. 63 (5) (1996) 797S–811S.

[5] M.B. Zimmermann, R.F. Hurrell, Nutritional iron deficiency, Lancet 370 (9586)(2007) 511–520.

[6] B. Halliwell, Biochemistry of oxidative stress, Biochem. Soc. Trans. 35 (Pt. 5)(2007) 1147–1150.

[7] B. Halliwell, J.M.C. Gutteridge, Free Radicals in Biology and Medicine, OxfordUniversity Press, Oxford/New York, 2007.

[8] G.D. Ginder, Microcytic and hemochromic anemias, in: L. Goldman, D.A.Aussiello, W. Arend, J.O. Armitage, D. Clemmons, J. Drazen, R. Griggs, J. Newman,E. Foster (Eds.), Cecil Medicine, Saunders Elsevier, Philadelphia, PA, 2008.

[9] H. Kodama, C. Fujisawa, Copper metabolism and inherited copper transport dis-orders: molecular mechanisms, screening, and treatment, Metallomics 1 (2009)42–52.

10] M.C Linder, L. Wooten, P. Cerveza, S. Cotton, R. Shulze, N. Lomeli, Copper trans-port, Am. J. Clin. Nutr. 67 (5 Suppl.) (1998) 965S–971S.

11] R.R. Crichton, S. Wilmet, R. Legssyer, R.J. Ward, Molecular and cellular mecha-nisms of iron homeostasis and toxicity in mammalian cells, J. Inorg. Biochem.91 (1) (2002) 9–18.

12] E. Madsen, J.D. Gitlin, Copper and iron disorders of the brain, Annu. Rev. Neu-rosci. 30 (2007) 317–337.

13] K.J. Barnham, A.I. Bush, Metals in Alzheimer’s and Parkinson’s diseases, Curr.Opin. Chem. Biol. 12 (2) (2008) 222–228.

14] X. Huang, R.D. Moir, R.E. Tanzi, A.I. Bush, J.T. Rogers, Redox-active metals, oxida-tive stress, and Alzheimer’s disease pathology, Ann. NY Acad. Sci. 1012 (2004)153–163.

15] M.E Gotz, K. Double, M. Gerlach, M.B. Youdim, P. Riederer, The relevance ofiron in the pathogenesis of Parkinson’s disease, Ann. NY Acad. Sci. 1012 (2004)193–208.

16] H. Hayashi, M. Yano, Y. Fujita, S. Wakusawa, Compound overload of copperand iron in patients with Wilson’s disease, Med. Mol. Morphol. 39 (3) (2006)121–126.

17] J. Aaseth, T.P. Flaten, O. Andersen, Hereditary iron and copper deposition: diag-nostics, pathogenesis and therapeutics, Scand. J. Gastroenterol. 42 (6) (2007)673–681.

gical

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

M.E. Letelier et al. / Chemico-Biolo

18] M.P Mattson, Metal-catalyzed disruption of membrane protein and lipid sig-naling in the pathogenesis of neurodegenerative disorders, Ann. NY Acad. Sci.1012 (2004) 37–50.

19] M.E. Letelier, A.M. Lepe, M. Faundez, J. Salazar, R. Marin, P. Aracena, H.Speisky, Possible mechanisms underlying copper-induced damage in biolog-ical membranes leading to cellular toxicity, Chem. Biol. Interact. 151 (2) (2005)71–82.

20] M.E. Letelier, M. Faúndez, J. Jara-Sandoval, A. Molina-Berríos, J. Cortés-Troncoso, P. Aracena-Parks, R. Marín-Catalán, Mechanisms underlying theinhibition of the cytochrome P450 system by copper ions, J. Appl. Toxicol. 135(3) (2010) 208–214.

21] M.E. Letelier, F. Lagos, M. Faundez, D. Miranda, M. Montoya, P. Aracena-Parks,V. Gonzalez-Lira, Copper modifies liver microsomal UDP-glucuronyltransferaseactivity through different and opposite mechanisms, Chem. Biol. Interact. 167(1) (2007) 1–11.

22] M.E. Letelier, M. Martinez, V. Gonzalez-Lira, M. Faundez, P. Aracena-Parks, Inhi-bition of cytosolic glutathione S-transferase activity from rat liver by copper,Chem. Biol. Interact. 164 (1–2) (2006) 39–48.

23] B. Halliwell, J.M. Gutteridge, Role of free radicals and catalytic metal ions inhuman disease: an overview, Methods Enzymol. 186 (1990) 1–85.

24] M.E. Letelier, A. Molina-Berrios, J. Cortes-Troncoso, J. Jara-Sandoval, A. Müller,P. Aracena-Parks, Comparative effects of superoxide anion and hydrogen per-oxide on microsomal and cytosolic glutathione S-transferase activities of ratliver, Biol. Trace Elem. Res. 134 (1) (2010) 203–211.

25] Y. Aniya, M.W. Anders, Activation of rat liver microsomal glutathione S-transferase by reduced oxygen species, J. Biol. Chem. 264 (4) (1989) 1998–2002.

[

[

Interactions 188 (2010) 220–227 227

26] Y. Aniya, M.W. Anders, Activation of rat liver microsomal glutathione S-transferase by hydrogen peroxide: role for protein-dimer formation, Arch.Biochem. Biophys. 296 (2) (1992) 611–616.

27] O.H. Lowry, N.J. Rosebrough, A.L. Farr, R.J. Randall, Protein measurement withthe Folin phenol reagent, J. Biol. Chem. 193 (1) (1951) 265–275.

28] M.E. Letelier, A. Pimentel, P. Pino, A.M. Lepe, M. Faundez, P. Aracena, H. Speisky,Microsomal UDP-glucuronyltransferase in rat liver: oxidative activation, BasicClin. Pharmacol. Toxicol. 96 (6) (2005) 480–486.

29] L.M. Gaetke, C.K. Chow, Copper toxicity, oxidative stress, and antioxidant nutri-ents, Toxicology 189 (1–2) (2003) 147–163.

30] S Altamura, M.U. Muckenthaler, Iron toxicity in diseases of aging: Alzheimer’sdisease, Parkinson’s disease and atherosclerosis, J. Alzheimers Dis. 16 (4) (2009)879–895.

31] J.M. Gutteridge, B. Halliwell, Iron toxicity and oxygen radicals, Baillieres Clin.Haematol. 2 (2) (1989) 195–256.

32] F. Pinnen, I. Cacciatore, C. Cornacchia, P. Sozio, L.S. Cerasa, A. Iannitelli, C. Nasuti,F. Cantalamessa, D. Sekar, R. Gabbianelli, M.L. Falcioni, A. Di Stefano, Codrugslinking L-dopa and sulfur-containing antioxidants: new pharmacological toolsagainst Parkinson’s disease, J. Med. Chem. 52 (2) (2009) 559–563.

33] C. Mancuso, T.E. Bates, D.A. Butterfield, S. Calafato, C. Cornelius, A. De Lorenzo,

A.T. Dinkova Kostova, V. Calabrese, Natural antioxidants in Alzheimer’s disease,Expert Opin. Investig. Drugs 16 (12) (2007) 1921–1931.

34] R.R. Crichton, R.J. Ward, Iron chelators and their therapeutic potential, Met. IonsBiol. Syst. 41 (2004) 185–219.

35] T. Wang, Z. Guo, Copper in medicine: homeostasis, chelation therapy and anti-tumor drug design, Curr. Med. Chem. 13 (5) (2006) 525–537.