Iron chelators increase the resistance of Ataxia telangeictasia cells to oxidative stress

DNA Repair 3 (2004) 1263–1272

Iron chelators increase the resistance ofAtaxia telangeictasiacells to oxidative stress

Rodney E. Shackelforda,b, Ryan P. Manuszaka, Cybele D. Johnsonb,Daniel J. Hellrunga, Charles J. Linka, Suming Wanga,∗

a Iowa Cancer Research Foundation, 11043 Aurora Avenue, Urbandale, IA 50322, USAb Osteopathic Medical Center, Des Moines University, Des Moines, IA 50309, USA

Accepted 12 January 2004

Available online 27 April 2004

Abstract

Ataxia telangeictasia(A-T) is an autosomal recessive disorder characterized by immune dysfunction, genomic instability, chronic oxidativedamage, and increased cancer incidence. Previously, desferal was found to increase the resistance of A-T, but not normal cells to exogenousoxidative stress in the colony forming-efficiency assay, suggesting that iron metabolism is dysregulated in A-T. Since desferal both chelatesiron and modulates gene expression, we tested the effects of apoferritin and the iron chelating flavonoid quercetin on A-T cell colony-formingability. We demonstrate that apoferritin and quercetin increase the ability of A-T cells to form colonies. We also show that labile iron levelsare significantly elevated in Atm-deficient mouse sera compared to syngeniec wild type mice. Our findings support a role for labile iron actingas a Fenton catalyst in A-T, contributing to the chronic oxidative stress seen in this disease. Our findings further suggest that iron chelatorsmight promote the survival of A-T cells and hence, individuals with A-T.© 2004 Elsevier B.V. All rights reserved.

Keywords: Ataxia telangeictasia; Labile ferrous iron; Iron chelator; Copper; Quercetin

1. Introduction

Ataxia telangeictasia(A-T) is an autosomal recessivedisorder characterized by immune dysfunction, prematureaging, increased ionizing radiation (IR) sensitivity, oculo-cutaneous telangiectasias, cerebellar degeneration accom-panied by ataxia, and increased lymphoreticular cancerincidence[1]. In culture cells from individuals with A-Tshow multiple defects when compared to normal cells,including increased serum growth factor requirements, in-creased genomic instability, premature senescence, andfailure to exhibit the G1, S, or G2 checkpoint responses,or p53 induction following IR[1]. The gene mutated inA-T, ATM (for A-T- mutated) has been cloned and its prod-

Abbreviations:A-T, Ataxia telangeictasia; ATM, A-T-mutated; FBS,fetal bovine serum; IR, ionizing radiation;t-BOOH, tert-butyl hydroperox-ide; BPS, bathophenantroline disulphonate; SOD, superoxide dismutase;DAPI, 4′,6-diamidino-2-phenylindole

∗ Corresponding author. Tel.:+1-515-241-8740; fax:+1-515-241-8788.E-mail address:[email protected] (S. Wang).

uct exhibits serine/theonine kinase activity induced by IR,and oxidants such ast-butyl hydroperoxide (t-BOOH),chromium VI, and nitric oxide[2–6]. ATM activation ap-pears to follow changes in chromatin structure resulting inthe dissociation of inactive ATM multimers into monomers,with concomitant activating autophosphorylation[7].

Evidence indicates that A-T is in part, a disease of in-creased oxidative stress, diminished antioxidant capacity,and inability to respond appropriately to exogenous ox-idants [8]. For example, A-T cells show increased lipidperoxidation, lowered catalase activity, lowered manganesesuperoxide dismutase levels, and delayed glutathione resyn-thesis after depletion with diethylpyrocarbonate[9,10]. Inculuture, A-T cells exhibit elevated levels of stress-inducedproteins which are lowered by treatment with the antioxidant�-lipoic acid, indicating that these cells are under chronicoxidative stress[11]. Individuals with A-T exhibit increasedmarkers of oxidative stress, including lipid peroxidaton and8-hydroxydeoxyguanosine[12]. Atm-deficient mice showa similar pathophysiology, particularly in the cerebellum[13,14]. In vitro, A-T cells are unusually sensitive to the

1568-7864/$ – see front matter © 2004 Elsevier B.V. All rights reserved.doi:10.1016/j.dnarep.2004.01.015

1264 R.E. Shackelford et al. / DNA Repair 3 (2004) 1263–1272

toxic effects of exogenous oxidants, including H2O2, t-BOOH, nitric oxide, chromium VI, arsenic, and superoxide,and fail to exhibit G1 and G2 checkpoint responses, or p53induction followingt-BOOH exposure[4–6,15–18].

Previously, the importance of oxidative stress in A-T androle of labile iron in mediating the toxic effects oft-BOOH[19], led us to hypothesize that desferal would increase theresistance of A-T cells to this oxidant[20]. We demonstratedthat desferal pretreatment increased the ability of A-T, butnot normal cells, to form colonies in the colony-efficiencyforming assay followingt-BOOH exposure[4]. Addition-ally, A-T cells exhibited increased sensitivity to the toxiceffect of FeCl3 in the colony forming-efficiency assay com-pared to normal cells and failed to exhibit an FeCl3-inducedG2 checkpoint response. The data suggested that labile ironis dysregulated in A-T and desferal increases the resistanceof A-T cells tot-BOOH via labile iron chelation. Such chela-tion would lessen Fenton chemistry and thereby increaseA-T cell survival and colony forming ability.

While deferal is a powerful iron chelator, it also modu-lates gene activity, such as hypoxia inducible factor[21],raising the possibility that altered gene expression andnot iron chelation explains desferal’s effect on A-T cells.To further examine iron chelation and labile iron in thepathobiochemistry of A-T, we examined the effect of twopharmacologically different iron chelators on A-T cells. Wereport that apoferritin and the flavonoid quercetin increasethe plating efficiency of A-T, but not normal cells in long-term culture. Apoferritin also increased the resistance ofA-T, but not normal cells, to the toxic effects oft-BOOH inthe colony forming-efficiency assay. Quercertin had similareffects on both normal and A-T cells, although to a fargreater extent in A-T cells. These later two observationsindicate that apoferritin and quercetin also increase the re-sistance of A-T cells to oxidative stress. Additionally wereport that labile iron is elevated in the sera of Atm-deficientmice compared to syngeneic normal mice. Taken together,our data indicates that iron chelation increases the resis-tance of A-T cells to oxidative stress, an effect probablydependent upon dysregulated iron metabolism in A-T.

2. Materials and methods

2.1. Materials

FeCl3, CuCl2, Pb acetate, NiCl2, (NH4)2Fe(SO4)2,CoCl2, bathophenantroline disulphonate (BPS), 4′,6-diamidino-2-phenylindole (DAPI), apoferritin, and hy-gromycin were obtained from Sigma Chemical Corp. (St.Louis, MO). FBS (fetal bovine serum) and Dulbecco’sModified Eagle’s Medium (DMEM) were obtained fromInvitrogen (Rockville, MD). Culture dishes were obtainedfrom Becton Dickinson (Franklin Lakes, NJ). The Hu-man Ferritin Quantification ELISA Kit (Cat. # 1810)was purchased from Alpha Diagnostic International, San

Antonio, TX and used according the manufacturer’sprotocol.

2.2. Cells

VA13, an SV40-immortalized normal fetal lung fibrob-last cell line, was obtained from ATCC (Rockville, MD,USA). AT22IJE-T (AT22) is an SV40-immortalized A-T fibroblast cell line. AT22IJE-T pEBS7 (pEBS7) andAT22IJE-T pEBS7-YZ5 (YZ5) cells are from the parentalAT22 cell line transduced with either a hygromycin re-sistance expression vector or the same vector with ATMexpression, respectively. These three AT22 cell lines weregenerous gifts from Dr. Yosif Shiloh at the Sacker Schoolof Medicine, Tel Aviv University, Israel[22]. The NHF1normal foreskin human fibroblast primary cell strain wasa generous gift from Dr. Richard Paules at the NationalInstitute of Environmental Health Sciences. The primarydermal A-T fibroblast strains ATDM-1 and ATDM-2 wereobtained by skin biopsy from two unrelated males, ages10 and 14, diagnosed with A-T and previously demon-strated to lack ATM expression[20]. The AT22 and VA13cell lines were cultured in DEMEM containing 5% FBSand 1% penicillin/streptomycin/glutamate. The pEBS7 andpEBS7-YZ5 cell lines were grown in the same media with100�g/ml hygromycin [22]. The primary fibroblast cellstrains were grown in DEMEM with 20% FBS and 1%penicillin/streptomycin/glutamine.

2.3. A-T and normal seras

Frozen seras from five individuals with A-T (ages 9, 11,13, 22, and 34) and four age-matched controls (ages 10, 21,22, and 36) were kindly provided by Dr. Howard Ledermanat the John Hopkins A-T Clinic.

2.4. Colony forming-efficiency assay

Colony forming-efficiency experiments were performedas previously described[4]. In brief, exponentially grow-ing cells were plated at 2000 cells/100 mm tissue culturedish in 10 ml appropriate media with an iron chelator andcultured for 14 days. The resulting colonies were fixedand stained by water:methanol addition (1:1) containingcrystal violet (1 g/L). “Colonies” consisted of cell clus-ters containing greater than 50 cells when counted undera dissecting microscope. For experiments involving apo-ferritin or quercetin pre-exposure ont-BOOH resistance,the cells were allowed to adhere 6 h in increasing apo-ferritin or quercetin concentrations, washed 4× with 5 mlmedia, treated with 10�M t-BOOH for 15 min in appro-priate media, washed 1 X with 5 ml media, and allowedto grow for 14 days in 10 ml of media. Data indicatessurvival as a percentage of untreated cells. The AT22 andVA13 immortalized fibroblast lines were used at passages20–40. The primary NHF1, ATDM-1, and ATDM-2 cell

R.E. Shackelford et al. / DNA Repair 3 (2004) 1263–1272 1265

strains were used at passages 6–19. All experiments weredone at least twice in triplicate. Standard deviations (errorbars) were calculated from each experimental data pointdivided by the mean untreated value and averaged betweentriplicate experiments to obtain the mean standard devia-tion.

2.5. Mitotic delay

Mitotic delay assays were performed as previoulsy de-scribed[4]. In brief, cells were plated onto 100 mm tissueculture plates and incubated 72 h (to approximately 50%confluency). The cells were treated with 100�M CuCl2, Pbacetate, NiCl2, or CoCl2 for 15 min, washed 2× with 5 mlmedia, cultured 2 h, and the media was removed. The cellswere fixed with 5 ml cold methanol for 10 min and air-dried.The cells were stained with 0.2�g/ml DAPI in water and ex-amined by fluorescence microscopy. The percentage of mi-totic cells (miotic index) was determined by counts of 5000cells. Mitotic delay experiments were performed in tripli-cate and the standard deviations calculated as in the colonyforming-efficiency assay.

2.6. Atm-deficient mouse status typing

Atm-heterozygous mice were obtained from The JacksonLaboratory (Bar Harbor, ME) and were cared for under ap-proval of an animal protocol using American Associationfor Laboratory Animal Science guidelines. ATM genotyp-ing was performed as previously described[23]. In brief,heterozygous mice pairs were bred and Atm-deficient andwild type mice were identified by extracting 35–70�l ofblood from the Saphenous vein with heparinzed microcap-illary tubes. The blood was transferred to a microfuge tubewith 20�l of 10 mM EDTA, mixed, and stored on ice. Twohundred�l lysis buffer (0.32 M Sucrose, 10 mM Tris–HClpH 7.5, 5 mM MgCl2, 1% v/v Triton X-100) was addedto each tube. The tubes were vortexed and centrifuged at16,000 g to pellet. The samples were washed 3× with ly-sis buffer and centrifuged as before. The pellets were re-supsended in 25–50�l of digestion buffer (50 mM KCl,10 mM Tris–HCl pH 8.3, 2.5 mM MgCl2, 0.1 mg/ml gelatin,0.45% [v/v] Nonidet P40, 0.45% [v/v] Tween 20). The so-lution was autoclaved to sterilize and dissolve the gelatin.Proteinase K, was added at a final working concentration of60�g/ml, to the digestion buffer immediately prior to use.The resulting solution was incubated at 55◦C overnight,heated to 97◦C for 10 min to inactivate the protenase K,and PCR was performed on 2�l of the samples using fourprimers designed to amplify either the Neo genetic mark-ers or the ATM gene. The PCR reactions were run into a1.5% argarose gel. The wild type mice had only one bandat 147 bp utilizing the A-T primers. The Atm-deficient micemice had only one band at 280 bp utilizing the Neo primers[23].

2.7. Atm-deficient and normal murine seras

The Atm-deficient and normal murine seras were obtainedvia cardiac puncture of deeply anesthesia mice which werethen sacrificed. The blood from each mouse was allowed toclot in a 1.5 ml eppindorf tube for 30 min at room temp. After30 min the blood was centrifuged with a quick spin (∼30 s)at 9000× g and the upper sera layer was removed. Thesera was immediately used to quantify labile ferrous iron.All mice were litermates between 20 and 35 days old andall mice were the same age when labile iron concentrationswere quantified. The sera from each mouse was measuredseparately. The seras of five normal and five Atm-deficientmice were quantified in two separate experiments.

2.8. Quantification of labile ferrous iron

The concentration of labile ferrous iron in normaland Atm-deficient mouse sera was quantified as previ-ously described[24]. In brief, a 50 mM stock solution ofbathophenantroline disulphonate (BPS) was prepared bydissolving 28.2 mg BPS in 1 ml of deionized pyrogen freewater. A standard curve for ferrous iron was constructedby dissolving 196 mg (NH4)2Fe(SO4)2 in 500 ml deionizedpyrogen free water. The solution was then diluted to make0.1–30�M (NH4)2Fe(SO4)2 solutions. Ten microliters ofBPS stock solution was added to 490�l of water or each(NH4)2Fe(SO4)2 solution, mixed, and left for 15 min to in-sure complete complex formation. Fifty microliters of eachsample was then transferred to a 96 well plate and the ab-sorbance was read at 535 nm against a 50�l water blank. Forserum labile ferrous iron measurements 49�l of each serasample received either 1�l of BPS stock solution or 1�lwater and was gently mixed a 96 plate well. After 15 min,the absorbance was read at 535 nm with the standard curve.The value of each respective blank was subtracted fromeach of the samples containing the BPS complex. Data wasanalyzed using a student’s one-tailedt-test demonstrating asignificant difference between the means of wild-type andAtm-deficient mouse sera (P = 0.013).

3. Results

3.1. Apopferritin and quercetin increase the platingefficiency of A-T, but not normal cells, in the colonyforming-efficiency assay

Previously we found that the iron chelator desferal in-creased the colony formation in A-T, but not normal cellsin the colony forming-efficiency assay[20]. While desferalchelates iron, it also modulates gene activity[21], raisingthe possibility that gene modulation and not iron chelationcaused the increased colony formation seen in A-T cells.To differentiate between these two possibilities we culturednormal and A-T cells for 14 days with either apoferritin


Fig. 1. (A) and (B) Long-term culture with apoferritin increases the platingefficiency of A-T cells in the colony forming-efficiency assay. (A) theeffect of increasing apoferritin concentrations upon two normal and threeA-T cell types was examined. (B) the effect of increasing apoferritinconcentrations upon the pEBS7 and pEBS7-YZ5 cell lines was examined.Exponentially growing cells were plated at a density of 2000 cells/100 mmplate in media with increasing apoferritin concentrations and cultured for14 days. Cell colonies were then stained and counted. Data indicatessurvival as a percentage of untreated cells.

or quercetin. Culturing the VA13, NHF1, AT22, ATDM-1,and ATDM-2 cells with increasing concentrations of apo-ferritin increased A-T cell colony formation, but not normalcell colony formation up to 30�g/ml apoferritin (Fig. 1A).In culture apoferritin is taken up by cells and binds labileiron, acting as an iron sequestrant and antioxidant[25,26].Thus apoferritin, a protein iron chelator pharmacologicallyunlike desferal, has similar effects to desferal on A-T colonyformation. Interestingly, apoferritn at higher concentrationswas more toxic to the immortalized cell lines than to theprimary cells (Fig. 1A).

To ascertain if this phenomenology was ATM depen-dent, similar experiments were carried out with the pEBS7

and pEBS7-YZ5 cell lines. The pEBS7 cell line carries ahygromycin resistance expression vector, while the pEBS7-YZ5 cell line carries a vector for pATM expression andhygromycin resistance. When the pEBS7 and pEBS7-YZ5cells lines were plated for 14 days with increasing apofer-ritin, colony formation increased only in the pEBS7 cells(Fig. 1B). High concentrations of apoferritin (100�g/ml)however, inhibited colony formation in both cell types(Fig. 1B). Thus, ATM expression in A-T cells blocks theincreased colony formation seen with apoferritin treatedA-T cells. As previously found[22], Western blot analysisdemonstrated that the pEBS7-YZ5 cells expressed ATM,while the pEBS7 cells did not (data not shown).

Quercetin is a flavonoid with multiple activities, includ-ing antioxidant and iron and copper chelating activities[27].Although not specifically an iron chelator, we examined theeffect of quercetin as it has low toxicity, powerful antiox-idant effects, and chelates iron, making it a possible phar-macological agent for treating A-T. As shown inFig. 2A,quercetin treatment resulted in a moderate, but significantincrease in the plating efficiency of A-T, but not normalcells, over a 1–10 nM range. Similar results were obtainedwith the pEBS7 cell line, but not the pEBS7-YZ5 cell line(Fig. 2B), indicating that this phenomenon was dependentupon a lack of ATM expression.

3.2. Apoferritin pretreatment increases the resistance ofA-T, but not normal cells, to the toxic effects t-BOOH inthe colony forming-efficiency assay

Our finding that apoferritin, quercetin, and desferal allincrease the plating efficiency of A-T, but not normal cells, inthe colony forming-efficiency assay demonstrates that ironchelators have colony formation-promoting effects specificto A-T cells[20]. These experiments do not demonstrate thatthese agents increase the resistance of A-T cells to oxidativestress, an important factor in the pathobiochemistry of A-T. To examine whether or not apoferritin could increasethe resistance of A-T cells to oxidative stress the VA13,AT22, NHF1, ATDM-1, and ATDM-2 cells were plated intomedia containing increasing apoferritin concentrations for 6h. The plates were washed 4× with media, the media wasreplaced, and the cells were treated 15 min with 10�M t-BOOH and allowed to form colonies. As shown inFig. 3,increasing apoferritin concentrations increased the colonyformation of A-T, but not normal cells in the colony forming-efficiency assay. We usedt-BOOH, as catalase activity hasbeen reported to be low in A-T cells andt-BOOH is poorlyhydrolyzed by catalase[9,28].

3.3. Quercetin increases the ability of A-T and normalcells to form colonies following t-BOOH exposure

To examine the effect of quercetin on resistance to oxida-tive stress in A-T and normal cells, the cells used inFig. 4were pretreated 6 h with increasing quercetin concentrations,


Fig. 2. (A) and (B) Long-term culture with quercetin increases the platingefficiency of A-T cells in the colony forming-efficiency assay. (A) theeffect of increasing quercetin concentrations upon two normal and threeA-T cell types was examined. (B) the effect of increasing quercetinconcentrations upon the pEBS7 and pEBS7-YZ5 cell lines was examined.Data indicates survival as a percentage of untreated cells.

exposed to 10�M t-BOOH, and colony formation wasquantified. As shown inFig. 4 quercetin concentrations upto 30 nM increased cellular resistance tot-BOOH in all celltypes. The percentage increase in colony formation wasmuch higher for the A-T cells than the normal cells.

3.4. A-T cells exhibit increased sensitivity to the toxiceffects of copper exposure

The observation that quercetin, a flavonoid that chelatesboth iron and copper[27], increased the plating efficiencyand ability of A-T cells to form colonies aftert-BOOHexposure (Figs. 2A, B, and 4) indicated that A-T cellsmight show increased sensitivity to copper exposure in thecolony forming-efficiency assay. As shown inFig. 5, A-T

Table 1pEBS7 (ATM-null) cells fail to exhibit a G2 checkpoint response comparedto pEBS7-YZ5 (ATM-expressing) cells

Cell type/treatment pEBS7 (%)± S.D. pEBS7-YZ5 (%)± S.D.

Untreated 100.00± 7.38 100.0± 3.35CuCl2 102.62± 3.47 32.90± 0.74Pb acetate 101.45± 10.67 40.0± 1.70NiCl2 101.45± 8.23 22.58± 5.08CoCl2 104.36± 3.06 35.73± 1.78

CuCl2, Pb acetate, NiCl2, and CoCl2 induce an ATM dependent G2checkpoint. pEBS7 and pEBS7-YZ5 cell lines were plated 72 h, treated15 min with 100�M CuCl2, Pb acetate, NiCl2, or CoCl2, washed twice,cultured for 2 h, fixed in cold methanol, and stained with DAPI. Thepercentage of mitotic cells was counted by fluorescence microscopy. Theresults are expressed as a precentage of the mitotic index of the mock-treated population.

cells exhibited an increased sensitivity to the toxic effectsof CuCl2 exposure when compared to normal cells. Inter-estingly, the immortalized cell lines were more sensitive tothe toxic effects of copper exposure than were the primarycells. However, within each cell type (i.e., immortalizedand primary cells), the A-T cells were more sensitive to thetoxic effect of copper.

3.5. CuCl2, Pb acetate, NiCl2, and CoCl2 induce an ATMdependent G2 checkpoint response

Our finding that A-T cells exhibit preferential sensitivityto the colony-inhibiting effects of CuCl2 exposure, com-bined with previous data demonstrating that FeCl3 inducedan ATM dependent G2 checkpoint[20], lead us to hypothe-size that other toxic metals would also elicit an ATM depen-dent G2 checkpoint response. To test this hypothesis pEBS7and pEBS7-YZ5 cells were exposed to 100�M CuCl2, Pbacetate, NiCl2, or CoCl2 for 15 min and the G2 checkpointresponse was examined 2 h later. As shown inTable 1,CuCl2, Pb acetate, NiCl2, and CoCl2 all induced an ATMdependent G2 checkpoint. Thus toxic metals other thaniron induce an ATM dependent G2 checkpoint response.Time course experiments demonstrated that the G2 check-point response was maximal at 2 h for each metal (data notshown).

3.6. Atm-deficient mice exhibit increased serum ferrouslabile iron compared to syngeneic normal mice

While quercetin, apoferritin, and desferal are structurallyquite different and have different pharmacological proper-ties, they all chelate iron. Based on this data, we hypoth-esized that labile iron would be elevated in Atm-deficientcompared to syngeneic normal mice. When the concentra-tion of plasma serum labile ferrous iron in Atm-deficientmice was compared to that of syngeniec normal mice, theAtm-deficient mice exhibited significantly elevated levels offerrous iron (Fig. 6).


Fig. 3. Pretreatment of A-T and normal cells with apoferritin results in increased resistance tot-BOOH toxicity in the colony forming-efficiency assay.Exponentially growing NHF1, VA13, ATDM-1, ATDM-2, and AT22 cells were plated for 6 h with increasing concentrations of apoferritin, washed 4×,treated with 10�M t-BOOH 15 min, washed 1×, cultured for 14 days, and stained. Data indicates survival as a percentage of untreated cells. Untreatedvalues were VA13 cells, 100± 7.36; AT22, 100± 9.94; NHF1, 100± 4.34; ATDM-1, 100± 12.46; ATDM-2, 100± 5.24.

3.7. Ferritin is not dysregulated in A-T cells

No differences in ferritin induction by 10�M FeCl3 inthe pEBS7 and pEBS7-YZ5 could be detected by enzymelinked immunosorbent assay (Human Ferritin QuantificationELISA Kit, data not shown). Similarly, ferritin levels in the

Fig. 4. Pretreatment of A-T and normal cells with quercetin results in increased resistance tot-BOOH toxicity in the colony forming-efficiency assay.Exponentially growing NHF1, VA13, ATDM-1, ATDM-2, and AT22 cells were plated for 6 h with increasing concentrations of quercetin, washed 4×,treated with 10�M t-BOOH 15 min, washed 1×, cultured for 14 days, and stained. Data indicates survival as a percentage of untreated cells. Untreatedvalues were VA13 cells, 100± 5.92; AT22, 100± 3.56; NHF1, 100± 2.92; ATDM-1, 100± 6.43; ATDM-2, 100± 6.34.

frozen seras from five individuals with A-T were not sig-nificantly different than ferritin levels in four age-matchedcontrol individuals. The ferritin levels in the five A-T serumsamples were also well within the normal range for serumferritin (data not shown). Based on this data we concludethat ferritin is not dysregulated in A-T.


Fig. 5. A-T cells exhibit increased sensitivity to CuCl2 in the colony forming-efficiency assay. Exponentially growing NHF1, VA13, ATDM-1, ATDM-2,and AT22 cells were plated for 6 h and treated 15 min with increasing concentrations of CuCl2, washed 2×, cultured 14 days, stained, and counted. Dataindicates survival as a percentage of untreated cells.

Fig. 6. A-T mice exhibit increase labile serum ferrous iron compared tosyngeneic normal mice. Serum labile ferrous iron in syngeneic normal andA-T mice were measured against a (NH4)2Fe(SO4)2 standard curve viacalometric assay[24]. Data was analyzed using a student’s one-tailedt-testdemonstrating a significant difference between the means of wild-typeand A-T mouse sera (P = 0.013).

4. Discussion

Previously we found that the iron chelator desferal in-creased the plating efficiency of A-T, but not normal cells, inthe colony forming-efficiency assay. Desferal also increasedthe resistance of A-T, but not normal cells, to the toxic effectsof t-BOOH in the colony forming-efficiency assay. Desferalhas multiple pharmacological effects including iron chela-tion and modulation of gene expression[21]. Thus genemodulation and not iron chelation could be responsible forthe effects of desferal on A-T cells. To address this issue wetreated normal and A-T cells with apoferritin and quercetin.Long-term culture with apoferritin and quercetin increasedcolony formation in A-T, but not normal cells, in the colonyforming-efficiency assay. Apoferritin pretreatment also in-creased the resistance of A-T, but not normal cells, to thetoxic effects oft-BOOH in the colony forming-efficiencyassay. Based on the similar effects of apoferritin, quercetin,and desferal on A-T cells and their different structures andpharmacological activities, we conclude that iron chelationand not the gene-modulating effects of desferal are respon-sible for the effect of these agents on A-T cells.

Iron is known to mediate the toxic effects of peroxidesvia events such as the Fenton reaction[29]. Based on thisand previous observations[20], we hypothesized that la-bile iron would be elevated in A-T. To investigate this wecompared syngeneic Atm-deficient and normal mice andfound significantly elevated labile ferrous iron in the seraof the Atm-deficient mice by calormetric assay[24]. SinceAtm-deficient mice share much of the same pathobiologyas individuals with A-T[13,14], we conclude it likely thatlabile iron is dyregulated in individuals with A-T. Althoughwe attempted to measure labile ferrous iron in the seras ofseveral normal individuals and individuals with A-T, our


attempts failed as only frozen seras were available and thecalormetric assay of Nilsson et al.[24] works only with freshserum.

Apoferritin and desferal have properties that make themdifficult to use as a pharmacological treatment for A-T. Toour knowledge, apoferritin-based therapy in not availableand desferal is limited in its usefulness due to its extremelyshort (10 min) plasma half life and oral ineffectiveness[30].For this reason we investigated the effects of the flavonoidquercetin on A-T cells. Quercetin chelates iron and copperand is a powerful antioxidant[27]. Flavonoids also havecomparatively long serum half lives, low toxicity, and epi-demiological data indicates that high flavonoid intake cor-relates with a reduction in diseases associated with chronicoxidative stress[31,32].

In the colony forming-efficiency assay quercetin increasedA-T cell resistance tot-BOOH as measured by the ability toform colonies (Fig. 4). Quercetin had a similar effect on nor-mal cells, although to a lesser degree. The reason for this dif-ference probably lies in the lower sensitivity of normal cellsto oxidative stress compared to A-T cells and iron chelat-ing effect of quercetin which appears to preferentially effectA-T cells (20, above). Analysis of efficacy of quercetin orother flavonoids as orally active pharmacological agents totreat A-T must await such analyses in Atm-deficient mice.Since quercetin is also a copper chelator, we examined therelative toxicity of copper exposure on A-T and normal cells(Fig. 5). Copper was preferentially toxic to A-T cells, in-dicating that a Fenton catalyst other than iron is toxic toA-T cells. However at high copper concentrations (30 and100�M) the VA13 cells exhibited greater colony formationinhibition than did the primary A-T cells (Fig. 5). This dataand the observation that the FeCl3 elicited G2 checkpointresponse is ATM dependent[20] led us to examine the ef-fects of other toxic metals on A-T cells. Here we demon-strate that CuCl2, Pb acetate, NiCl2, and CoCl2 also inducean ATM dependent G2 checkpoint (Table 1). Thus, ATM isrequired for the G2 checkpoint following exposure to toxiclevel of multiple metals.

Interestingly, we observed a large amount of responsevariation between cell types used here. Currently we haveno explaination for this variation. However, we found twogeneral patterns of cellular behavior. First, the immortalizedcells lines exhibited greater sensitivity to oxidative stressthan did the corresponding wild type cells (i.e., AT22 vs.ATDM-1 and ATDM-2, and VA13 compared NHF1 cells).Also the VA13 cells exhibited greater colony formation in-hibition at higher copper concentrations than was observedwith the primary A-T cells (Fig. 5). Second, higher cellularsensitivity tot-BOOH correlated with a greater response ironchelators. Thus the AT22 cells, which exhibited high sen-tivity to t-BOOH (Figs. 3 and 4), also exhibited the greatestresponse to iron chelators (Figs. 1–4). The molecular basisof this correlation is poorly understood, but appears to in-volve an increased sensitivity to oxidative stress that followscellular immortalization[34].

The molecular mechanism describing the effect of ironchelators on A-T could involve several factors. One of themcould be dysregulated ferritin induction comparable to thep53 dysregulation seen in A-T[1]. We think this is unlikelyas in our hands induction of ferritin by FeCl3 was the samein pEBS7 and pEBS7-YZ5 cell lines and the sera of fiveindividuals with A-T did not show altered ferritin levels.Additionally, the relatively modest increase in labile ferrousiron seen in the plasma of the A-T mice (∼1.05�M inAtm-deficient versus 50 nM in wild type mice), is too lowto significantly impinge upon ferritin regulation. Anotherpossibility is that while ferritin levels are comparativelynormal in A-T, the chronic oxidative stress seen in A-T[11,12] may cause the release of labile iron from ferritin ashas been previously seen for superoxide treated ferritin[33].

A simple model of the possible biochemical defect in A-Tis shown inFig. 7. Mitochondrial superoxide is catalyzedby superoxide dismutase to produce H2O2. Since catalaseactivity is often lower in A-T cells[9], this could lead toa relative increase in the H2O2 pool. H2O2 in the presenceof the Fenton catalyst ferrous iron, would be converted tothe highly toxic hydroxyl radical, causing cellular damage.Superoxide could also act to reduce ferric iron back to itferrous form, thus allowing for a continuous supply of cat-alyst. Thus a relatively small amount of labile iron couldhave a significant biological effect. Iron chelators wouldsequester labile iron and thereby block Fenton chemistry.

Fig. 7. Model of possible ferrous iron activity and subsequent redoxreactions in A-T cells. Mitochondrial function produces superoxide, whichbecomes H2O2 via superoxide dismutase (SOD) activity. Lowered catalaseactivity (often seen in A-T[9]), results in an increased pool of H2O2,which in turn becomes the toxic hydroxyl radical via the presence offerrous iron acting as a Fenton catalyst (bold arrow). The ferric iron thatresults from the Fenton chemistry may be regenerated to ferrous iron viareduction from superoxide.


Support for this model comes from the previous observa-tion that increased Cu/Zn-superoxide dismutase expressionin Atm-deficient mice actually exacerbated radiation sen-sitivity and hematopoietic abnormalities[35]. Lastly, irontoxity has been linked to several neurogenerative disorders,including Hallervorden–Spatz syndrome and Parkinson’s,Alzheimer’s, and Huntington’s diseases, raising the possi-bility that iron chelation may be used to treat these diseases[36]. Since oxidative stress plays an important role in Purk-inje cell death in Atm-deficient mice[37] and labile ironis dysregulation in these mice, our data also suggests thatblood-brain barrier crossing iron chelators may slow Purk-inje cell death in A-T.

Acknowledgements

This research was partially supported by a grant from theA-T Children’s Project. Special thanks goes to Dr. HowardLederman of the John HopkinsAtaxia TelangiectasiaClinicfor providing us with A-T and normal human seras.

References

[1] M.F. Lavin, Y. Shiloh, The genetic defect inAtaxia telangiectasia,Annu. Rev. Immunol. 15 (1997) 177–202.

[2] S. Banin, L. Moyal, S. Shieh, Y. Taya, C.W. Anderson, L. Chessa,N.I. Smorodinsky, C. Prives, Y. Reiss, Y. Shiloh, Y. Ziv, Enhancedphosphorylation of p53 by ATM in response to DNA damage, Science281 (1998) 1674–1677.

[3] C.E. Canman, D.S. Lim, K.A. Cimprich, Y. Taya, K. Tamai, K.Sakaguchi, E. Appella, M.B. Kastan, J.D. Siliciano, Activation ofthe ATM kinase by ionizing radiation and phosphorylation of p53,Science 281 (1998) 1677–1679.

[4] R.E. Shackelford, C.L. Innes, S.O. Seiber, S.A. Leadon, A.N. Hein-loth, R.S. Paules, TheAtaxia telangiectasiagene product is requiredfor oxidative stress-induced G1 and G2 checkpoint function in hu-man fibroblasts, J. Biol. Chem. 276 (2001) 21951–21959.

[5] L. Ha, S. Ceryak, S.R. Patierno, Chromium (VI) activatesAtaxiaTelangiectasiamutated (ATM) protein. Requirement of ATM for bothapoptosis and recovery from terminal growth arrest, J. Biol. Chem.278 (2003) 17885–17894.

[6] L.J. Hofseth, S. Saito, S.P. Hussain, M.G. Espey, K.M. Miranda, Y.Araki, C. Jhappan, Y. Higashimoto, P. He, S.P. Linke, M.M. Quezado,I. Zurer, V. Rotter, D.A. Wink, E. Appella, C.C. Harris, Nitric oxide-induced cellular stress and p53 activation in chronic inflammation,Proc. Natl. Acad. Sci. USA 100 (2003) 143–148.

[7] C.J. Bakkenist, M.B. Kastan, DNA damage activates ATM throughintermolecular autophosphorylation and dimer dissociation, Nature421 (2003) 499–506.

[8] A. Barzilai, G. Rotman, Y. Shiloh, ATM deficiency and oxidativestress: a new dimension of defective response to DNA damage, DNARepair 1 (2002) 3–25.

[9] D. Watters, P. Kedar, K. Spring, J. Bjorkman, P. Chen, M. Gate, G.Birrell, B. Garrone, P. Srinivasa, D.I. Crane, M.F. Lavin, Localizationof a portion of extranuclear ATM to peroxisomes, J. Biol. Chem.274 (1999) 34277–34282.

[10] M.J. Meredith, M.L. Dodson, Impaired glutathione biosynthesis incultured humanAtaxia telangiectasiacells, Cancer Res. 47 (1987)4576–4581.

[11] M. Gatei, D. Shkedy, K.K. Khanna, T. Uziel, Y. Shiloh, T.K. Pan-dita, M.F. Lavin, G. Rotman,Ataxia telangiectasia: chronic activa-tion of damage-responsive functions is reduced by alpha-lipoic acid,Oncogene 20 (2001) 289–294.

[12] J. Reichenback, R. Schubert, D. Schindler, K. Muller, H. Bohles, S.Zielen, Elevated oxidative stress in patients withAtaxia telangiecta-sia, Antioxidants Redox. Sig. 4 (2002) 465–469.

[13] A. Kamsler, D. Daily, A. Hochman, N. Stern, Y. Shiloh, G. Rot-man, A. Barzilai, Increased oxidative stress inAtaxia telangiectasiaevidenced by alterations in redox state of brains from Atm-deficientmice, Cancer Res. 61 (2001) 1849–1854.

[14] Y. Ziv, A. Bar-Shira, I. Pecker, P. Russell, T. Jorgensen, I. Tsarfati,Y. Shiloh, Recombinant ATM protein complements the cellular A-Tphenotype, Oncogene 15 (1997) 159–167.

[15] M.H.L. Green, A.J. Marcovitch, S.A. Harcourt, J.E. Lowe, I.C. Green,C.F. Arlett, Hypersensitivity ofAtaxia telangiectasiafibroblasts to anitric oxide donor, Free Radic. Biol. Med. 22 (1997) 343–347.

[16] M. Vuillaume, M. Best-Belpomme, R. Lafont, M. Hubert, Y. De-croix, A. Sarasin, Stimulated production of ATP by H2O2 dispro-portionation in extracts from normal and xeroderma pigmentosumskins, and from normal,Ataxia telangiectasiaand simian virus 40transformed cell lines, Carcinogenesis 10 (1989) 1375–1381.

[17] A.J. Ward, P.L. Olive, A.H. Burr A, M.P. Rosin, Response of fibrob-last cultures fromAtaxia telangiectasiapatients to reactive oxygenspecies generated during inflammatory reactions, Environ. Mol. Mu-tagen. 24 (1994) 103–111.

[18] D. Menendez, G. Mora, A.M. Salazar, P. Ostrosky-Wegman, ATMstatus confers sensitivity to arsenic cytotoxic effects, Mutagenesis 16(2001) 443–448.

[19] A. Barbouti, P.T. Doulias, B.Z. Zhu, B. Frei, D. Galaris, Intracellulariron, but not copper, plays a critical role in hydrogen peroxide-induced DNA damage, Free Radic. Biol. Med. 31 (2001) 490–498.

[20] R.E. Shackelford, R.P. Manuszak, C.D. Johnson, D.J. Hellrung, T.A.Steele, C.J. Link, S. Wang, Desferrioxamine treatment increasesthe genomic stability ofAtaxia telangiectasia, Cells DNA Repair 2(2003) 971–981.

[21] D.M. Templeton, Y. Liu, Genetic regulation of cell function in re-sponse to iron overload or chelation, Biochim. Biophys. Acta 1619(2003) 113–124.

[22] Y. Ziv, A. Bar-Shira, I. Pecker, P. Russell, T. Jorgensen, I. Tsarfati,andY. Shiloh, Recombinant ATM protein complements the cellularA-T phenotype, Oncogene 15 (1997) 159–167.

[23] JAX®Mice Protocol Index [http://aretha]jax.org/pub-cgi/protocols/protocols.sh?objtype=protocol&protocolid=220.

[24] U.A. Nilsson, M. Bassen, K. Savman, I. Kjellmer, A simple andrapid method for the determination of “free” iron in biological fluids,Free Radic. Res. 36 (2002) 677–684.

[25] G. Balla, H.S. Jacob, J. Balla, M. Rosenberg, K. Nath, F. Apple,J.W. Eaton, G.M. Vercellotti, Ferritin: a cytoprotective antioxidantstrategem of endothelium, J. Biol. Chem. 267 (1992) 18148–18153.

[26] M.B. Juckett, J. Balla, G. Balla, J. Jessurun, H.S. Jacob, G.M. Ver-cellotti, Ferritin protects endothelial cells from oxidized low densitylipoprotein in vitro, Am. J. Pathol. 147 (1995) 782–789.

[27] L. Mira, M.T. Fernandez, M. Santos, R. Rocha, M.H. Florencio,K.R. Jennings, Interactions of flavonoids with iron and copper ions: amechanism for their antioxidant activity, Free Radic. Res. 36 (2002)1199–1208.

[28] G.W. Winston, W. Harvey, L. Berl, A.I. Cederbaum, The generationof hydroxyl and alkoxyl radicals from the interaction of ferrousbipyridyl with peroxides, Biochem. J. 216 (1983) 415–421.

[29] A. Barbouti, P.T. Doulias, B.Z. Zhu, B. Frei, D. Galaris, Intracellulariron, but not copper, plays a critical role in hydrogen peroxide-induced DNA damage, Free Radic. Biol. Med. 31 (2001) 490–498.

[30] E.M. Calleja, J.Y. Shen, M. Lesser, R.W. Grady, M.I. New, P.J. Giar-dina, Survival and morbidity in transfusion-dependent thalassemic pa-tients on subcutaneous desferrioxamine chelation, nearly two decadesof experience, Ann. N.Y. Acad. Sci. 850 (1998) 469–470.

http://arethajax.org/pub-cgi/protocols/

http://arethajax.org/pub-cgi/protocols/


[31] J.F. Young, L.O. Dragstedt, J. Haraldsdottir, B. Daneshvar, M.A.Kal, S. Loft, L. Nilsson, S.E. Nielsen, B. Mayer, L.H. Skibsted,T. Huynh-Ba, A. Hermetter, B. Sandstrom, Green tea extract onlyaffects markers of oxidative status postprandially: lasting antioxidanteffect of flavonoid-free diet, Br. J. Nutr. 87 (2002) 343–355.

[32] P. Knekt, J. Kumpulainen, R. Jarvinen, H. Rissanen, M. Helio-vaara, A. Reunanen, T. Hakulinen, A. Aromaa, Flavonoid intakeand risk of chronic diseases, Am. J. Clin. Nutr. 76 (2002) 560–568.

[33] R. Agrawal, P.K. Sharma, G.S. Rao, Release of iron from ferritinby metabolites of benzene and superoxide radical generating agents,Toxicology 168 (2001) 223–230.

[34] G. Loo, Redox-sensitive mechanisms of phytochemical-mediated in-hibition of cancer cell proliferation, J. Nutr. Biochem. 14 (2003)64–73.

[35] Y. Peter, G. Rotman, J. Lotem, A. Elson, Y. Shiloh, Y. Groner, Ele-vated Cu/Zn-SOD exacerbates radiation sensitivity and hematopoieticabnormalities of Atm-deficient mice, EMBO J. 20 (2001) 1538–1546.

[36] C.W. Levenson, Iron and Parkinson’s disease: chelators to the rescue?Nutr. Rev. 61 (2003) 311–333.

[37] P. Chen, C. Peng, J. Luff, K. Spring, D. Watters, S. Bottle, S. Furuya,M.F. Lavin, Oxidative stress is responsible for deficient survivaland dendritogenesis in Purkinje neurons fromAtaxia telangiectasiamutated mutant mice, J. Neurosci. 23 (2003) 11453–11460.

Iron chelators increase the resistance of Ataxia telangeictasia cells to oxidative stress

Documents

Transcript of Iron chelators increase the resistance of Ataxia telangeictasia cells to oxidative stress