Depletion of cellular iron by BPS and ascorbate: Effect on toxicity of adriamycin

Free RadicalBiology& Medicine,Vol. 20, No. 3, pp. 319 329, 1996 Copyright © 1996 ElsevierScience Inc. Printed in the USA. All rights reserved 0891-5849/96 $15.00 + .00 ELSEVIER

SSDI 0891-5849(96)02054-3

Original Contribution DEPLETION

OF

CELLULAR

EFFECT

ON

IRON

TOXICITY

BY OF

BPS

AND

ASCORBATE:

ADRIAMYCIN

SREEDEVI NYAYAPATI, GUL AFSHAN, FRANK LORNITZO, ROBERT W. BYRNES, and DAVID H. PETERING Department of Chemistry, University of Wisconsin-Milwaukee, Milwaukee, WI, USA

(Received 27 December 1994; Revised 3 August 1995; Accepted 17 August 1995) A b s t r a c t - - A new method was developed that reduces the intracellular iron content of cells grown in serumcontaining culture without involving the significant uptake of iron-chelating agents into cells. Negatively charged bathophenanthrolinedisulfonate (BPS), together with ascorbate, caused cells to lose much of their cellular iron without causing much depression in HL-60 or H9c2 (2-1) cell proliferation over a 48-h period. When added to serum supplemented RPMI-1640 culture media, BPS and ascorbate efficiently reduced and competed for iron in Fe(III) transferrin to form Fe(II)(BPS)3. The reaction also occurred with purified human iron-transferrin. When cells were incubated with growth medium containing serum that had been treated with BPS and ascorbate for 24 h, little or no BPS 2 or Fe(II)(BPS)34- entered the cells, according to direct measurements and in agreement with the highly unfavorable 1-octanol/water partition coefficients for these molecules. However, iron was mobilized out of both cell types. After 24 h incubation of cells in this medium, there was no change in the activities of catalase and superoxide dismutase, or in the concentration of glutathione. Glutathione peroxidase was elevated 9%. Using HL-60 and H9c2 (2-1) cells made iron deficient with BPS and ascorbate, HL-60 cells grown in definedgrowth media in the absence of iron-pyridoxal isonicotinoyl hydrazone, or Euglena gracilis cells maintained in a defined medium that was rigorously depleted of iron, it was shown that the cytotoxicity of adriamycin is markedly dependent on the presence of iron in each type of cell. Similar results were obtained when HL-60 cells were grown in RPMI-1640 culture medium and serum that had been incubated for 24 h in BPS and ascorbate and then chromatographed over a Bio-Rad desalting column to remove small molecules including BPS, ascorbate, and Fe(II)(BPS)3. Keywords--Iron, Cells, Doxorubicin, Adriamycin, Transferrin, Antioxidants, Chelators, Iron deficient, Free radicals

these agents are thought to enter cells, they m a y interact directly with intracellular p o o l s o f the metal. 3"40P has substantial affinity for Fe 2+, Zn 2+, and both oxidation states o f Cu, bringing into question its specificity for iron. 5 Furthermore, O P is highly toxic to cells. 3 Similarly, although D F O exhibits a particularly large affinity for F e 3+, it also binds Cu 2+ with a sizable apparent stability constant, 6 and it, too, displays s o m e toxicity t o w a r d cells. 7 The p o s s i b i l i t y that both agents m a y interact with a variety o f m e t a l - b i n d i n g structures and p o s s i b l y other nonmetal sites c o m p l i c a t e s the interpretation o f their m e c h a n i s m s o f action. T w o other systems that have been explored recently are Euglena gracilis cells grown in a defined culture m e d i u m that has been r i g o r o u s l y depleted o f iron, and human H L - 6 0 cells adapted to g r o w in defined growth

INTRODUCTION M e t h o d s that reduce the concentration o f iron available to cells in culture or that l o w e r intracellular iron p o o l s will aid researchers w h o are a) studying the m e c h a nisms o f regulation o f proteins i n v o l v e d in iron transport, storage, and m e t a b o l i s m , b) those interested in the effects o f iron deficiency on cells, and c) others studying the role o f cellular iron in processes o f o x i d a n t d a m a g e . C o m m o n l y , metal chelating agents such as 1,10-phenanthroline (OP) or d e s f e r r i o x a m i n e ( D F O ) have been a d d e d to s c a v e n g e a v a i l a b l e iron. ~.2 B e c a u s e

Address correspondence to: David H. Petering, Department of Chemistry, University of Wisconsin-Milwaukee, P.O. Box 413, Milwaukee, WI 53201, USA. 319

320

S. NYAVAPA'N et al.

medium in the absence of iron-transferrin, which employs iron-pyridoxal isonicotinoyl hydrazone as the adjustable source of iron in place of iron-transferrin. ~ In each system, cellular iron deficiency resulted from the lack of a nutrient source of iron. A more general method to deprive cells of nutrient iron is needed for cells not adapted to growth in defined media in the absence of serum or iron-transferrin. The present communication describes a procedure that appears to reduce cellular iron by limiting access to serum iron without affecting the intracellular concentration of Zn 2 ' or causing overt toxicity at concentrations similar to those used to deplete cells of iron. In order to test the applicability of this new model for iron-deficient cells for the study of drugs that may use iron in their mechanism of action, the requirement of cellular iron for the cytotoxic properties of adriamycin has been examined. Adriamycin has a rich oxidation-reduction chemistry that can be coupled to the production of hydroxyl radicals through iron or copperdependent Fenton chemistry. 9.n~ It can also bind iron directly and participate directly in iron-dependent redox processes. ~ ~4Alternatively, adriamycin is a topoisomerase II inhibitor that can produce protein-associated DNA strand breakage.~S No iron requirement has been identified for this process. In the work described in this article, cells have been depleted of iron by three different means and then examined for their sensitivity to adriamycin.

METHODS AND MATERIALS

Materials RPMI- 1640, Dulbecco' s Modified Eagle' s Medium, trypsin-EDTA solution (10X), trypsin (powder form), human iron saturated transferrin, Hepes, sodium salt, Trizma, nitro blue tetrazolium, xanthine, xanthine oxidase, catalase, diethylenetriaminepentaacetic acid, potassium phosphate, sulfosalicylic acid, 2,6-dichlorophenol-indophenol, trisodium citrate, p-chloromercuribenzoic acid, glutathione, and adriamycin were purchased from Sigma Chemical Company (St. Louis, MO). Glutathione disulfide, bathophenanthroline disulfonate, ascorbate, vinylpyridine, and 1,10-phenanthroline were purchased from Aldrich Chemicals (Milwaukee, WI). Hydrogen peroxide and l-octanol were purchased from Fisher (St. Louis, MO). Acetic acid and phosphoric acid were obtained from J. T. Baker (Phillipsburg, NJ). Sodium cyanide was obtained from Mallinkrodt. These reagents were used without any further purification. All the solutions were made in doubly purified water, except for trypsin (powder form), which was dissolved in phosphate-buffered saline. Fe(II)BPSs

was prepared by adding BPS to aqueous ferrous ammonium sulfate at a 3:1 molar ratio. Human promyelocytic leukemia HL-60 cells and rat heart myoblasts, H9c2 (2-1), were purchased from American Type Culture Collection (Rockville, MD). HL-60 cells were grown in RPMI-1640 and the heart cells in Dulbecco's modified Eagle's medium. Both media were supplemented with 10% (v/v) fetal bovine serum. The cells were incubated at 37°C in an atmosphere of 5% CO> HL-60 cells were grown in suspension at an initial concentration of 7.0 x l0 s cells/ml, and H9c2 (2-1) cells were grown as monolayers at a starting cell density of 5 x 10s cells/ml.

Euglena gracilis. Euglena gracilis (Z Strain) was obtained from Carolina Biochemicals or the University of Texas. The cells were grown in a defined medium modified as described previously. ~6 Double-distilled, deionized water was used throughout. Stock solutions and preparations were held in glass containers previously treated with 10% analytical reagent nitric acid because traces of iron were observed to leach from various types of plastic. Growth medium containing less than 10 ppb Fe was made by passing medium without its normal complement of Fe through a "sandwich" of Mg(OH)2 as previously described. ~

Measurement of intracellular iron distribution. Total cellular iron and cytosolic iron were measured. HL-60 or H9c2 (2-1) cells were incubated in medium containing 10% fetal bovine serum until they reached maximal growth levels (about 2 - 3 x 106 cells/ml of HL-60 cells and near confluency, 4 - 5 X 106 cells/100 mm plate of monolayer H9c2 (2-1) cells) and then were treated without BPS, with 80/zM BPS, or with 80/zM BPS and 250 #M ascorbate for 24 h at 37°C in 5% CO2. After 24 h, the medium was discarded and the cells washed three times with cold PBS. H9c2 (2-1) monolayers were subsequently incubated with 1 ml of a 0.2% trypsin solution for 1-2 min, pipetted into a centrifuge tube, and centrifuged to obtain a pellet. For cytosolic iron measurement, cells were resuspended in 0.5 ml of water and 5 /.tL of 10% 2-mercaptoethanol and sonicated on ice with a Branson Cell Disrupter Model 200 (standard tip, 40% duty cycle, output level 5, for 60 pulses). The cell suspension was centrifuged at 48,000 x g for 20 min and the supernatant collected, applied to a calibrated Sephadex G-75 chromatographic column, and eluted with 20 mM Tris, pH 7.8. The fractions were analyzed for iron by graphite furnace atomic absorption on a GBC 904 AA spectrophotometer, and zinc by flame atomic absorption spectrophotometry using an Instrumentation Laboratory, Inc. (Lexington, MA) Model 357 atomic absorption spectrophotometer calibrated using a NBS-traceable zinc stan-

Cell iron deficiency and adriamycin toxicity dard solution. For determination of total cellular iron, cells washed with PBS were centrifuged and the pellets dissolved in 0.2 ml of hydrolysis solution (HC104 50%, HNO3 30%, H20 20%) and hydrolyzed over a steam bath for 30 min. Hydrolysis solution heated for 30 min in the same way was used as a blank. Both the blank and the sample solutions were diluted with double-distilled H20 and analyzed for iron and zinc by atomic absorption spectrophotometry. The number of cells used in each of the two experiments were 1 x 108 HL-60 cells and 5 x 107 H9c2 (2-1) cells.

Determination of catalase, superoxide dismutase (SOD), glutathione peroxidase, and glutathione (GSH + GSSG). In each experiment, two HL-60 cell cultures containing 7 × l0 s cells/ml were inoculated 48 h prior to assay. Twenty-four hours after inoculation, 144 #M BPS and 400 #M ascorbate were added to each culture and the cultures were allowed to grow an additional 24 h. The cultures were then pelleted by centrifugation, resuspended in phosphate-buffered saline (PBS), and centrifuged again. For catalase, SOD, and glutathione peroxidase assays, cells were resuspended in 1 ml PBS at 5 × 10 7 cells/ml. The suspensions were sonicated as those for iron and zinc measurements. The suspensions were centrifuged at 12,000 × g for 10 rain and the supernatants were assayed for enzyme activity. Total SOD activities were determined by the ability of aliquots to inhibit reduction of nitro blue tetrazolium (NBT) in the presence of xanthine, oxygen, and xanthine oxidase. ~7 Aliquots of cell extracts were added to a reaction mixture in a thermostatted cuvette (25°C) and the absorbance at 560 nm followed after the addition of sufficient xanthine oxidase, which produced an increase of about 0.025 absorbance units per minute in the absence of added extract. The final concentrations in the cuvette were xanthine, 0.15 mM; NBT, 56 #M; catalase, 1 U/ml; diethylenetriaminepenta-acetic acid, 1 mM; and 50 mM potassium phosphate buffer, pH 7.8. One unit of SOD is defined as the number of cells required to inhibit the SOD-inhibitable rate of increase in absorbance by 50%. Catalase activity was measured by following the decrease in H202 absorbance at 240 nm at 25°C. ~s The results are expressed in terms of first-order rate constants determined over a 15 s period, using an extinction coefficient of 40 M-~cm ~. Glutathione peroxidase was measured according to Geiger et al. in the presence of 0.2 mM t-butylhydroperoxide, 3 mM GSH, (I.2 mM NADPH, and 1.5 U/ ml glutathione reductase. L9 Aliquots of cell extracts were added to a thermostatted cuvette (37 °) containing these reagents and the absorbance decrease at 340 nm

321

due to NADPH utilization recorded. One unit of enzyme is defined as the amount of enzyme catalyzing oxidation of 1 nmol NADPH per minute. For measurement of GSH, the cell pellet was resuspended in 0.5 ml water, followed by acid extraction using acetic acid and sulfosalicylic acid. TM The extract was assayed for GSH and GSSG as described. 2°

Ascorbic acid measurement. The ascorbic acid content of cells was assayed by its ability to bleach the absorbance of 2,6-dichlorophenol-indophenol at 520 nm. 2~ Cells were passaged and grown 48 h prior to harvest with or without freshly prepared ascorbic acid during the final 24 h growth period. The cells were then washed as in the SOD and catalase assays. A cell extract was prepared by suspending 1 x 108 cells in cold 10% phosphoric acid and briefly vortexing the suspension. After centrifugation for 20 min at 3500 × g at 4°C, the supernatant was assayed in citrate/acetate buffer at pH 5.15 in the presence ofp-chloromercuriben-zoic acid to minimize interference from cellular thiols. 2~

Growth inhibition experiments The effect of various agents on HL-60 cell survival was examined in 24-well tissue culture plates. Stock cultures grown in RPMI- 1640 medium containing 10% fetal bovine serum were resuspended in fresh medium at 1.0 X 1 0 6 cells/ml and 1 ml of cell suspension was pipetted into each well. Then, various concentrations of BPS, ascorbate, BPS + ascorbate, Fe(II)(BPS)3, or OP were added to the wells. Each concentration was repeated in triplicate. Cells were incubated at 37°C in 5% CO2 for 48 h and the live cells counted using trypan blue dye exclusion as the criterion for viability. The percent of control cell survival (average of three experiments) was calculated as the ratio of treated cells to the control cells × 100.

Growth inhibition by adriamycin. H9c2 (2-1) cells were incubated in medium for 24 h at 37°C in 5% CO2, with or without 80 #M BPS and 250 #M ascorbate. After 24 h, the growth medium was removed and each plate of cells was incubated with 1 ml of 0.2% Sigma 10× trypsin at room temperature for 1 - 2 min. Once the cells were freed from the plates, they were centrifuged and the supernatant discarded. Then the cells were resuspended in the medium collected from the plates at 2.0 × 105 cells/ml. Finally, the cells (2 ml) were pipetted into sterile 35 × 10 mm culture dishes and an aliquot of the stock 10 mM adriamycin solution was added to each dish. Each concentration of drug was tested in triplicate. The cells were incu-

322

S. NYAYAPATIet al.

bated for 48 h at 37°C in 5% CO2. After 48 h, the medium was discarded and 0.1 ml of trypsin solution was added to each plate and incubated as above. After adding trypan blue, live cells were counted on a hemocytometer. The percentage of control cell survival was calculated as the ratio of treated cells to control cells (average of three experiments) × 100. HL-60 cells were incubated in 30 ml tissue culture flasks in RPMI-1640 medium supplemented with 10% fetal bovine serum in the presence or absence of 80 #M BPS and 250 #M ascorbate at 37°C, 5% CO2 for 24 h. After 24 h, the cells were centrifuged and the medium was collected into a sterile tissue culture flask. The cells were resuspended in the same medium at 1 × 106 cells/ml. One milliliter of the cells was pipetted into each well of a 24-well tissue culture plate and treated with adriamycin as described above. Alternatively, fetal bovine serum was incubated with or without 0.8 mM BPS and 2.5 mM ascorbate for 24 h and chromatographed over a Bio-Rad Econo Pac 10DG desalting column equilibrated with 10 mM PBS. The large molecular weight fractions eluted with PBS were pooled, filter sterilized, and used in place of regular serum. HL-60 cells were grown in this medium for 24 h. After 24 h, the cells were centrifuged and resuspended in the same medium at 1 × 106 cells/ ml and 1 ml was pipetted into each well of a 24-well tissue culture plate. To these cells iron-transferrin was added to give a final concentration of 5 /zg/ml. The cells were then incubated with adriamycin. HL-60 cells that have been adapted to grow in RPMI-1640 medium supplemented with iron-pyridoxal isonicotinoyl hydrazone in place of serum as the source of iron 8 were also treated with adriamycin. Cells were grown in medium with or without Fe-PIH for 48 h. After 48 h, the cells were centrifuged and resuspended in the same medium at 1 × 106 cells/ml and exposed to adriamycin. Heterotrophic Euglena gracilis cells were adapted to grow in a defined medium that contained a basal concentration of less than 10 nM Fe. Samples of Euglena gracilis were then grown in the basal medium to which was added 0, 9, 19, 36, or 180 nM iron for 2 weeks. Finally, 2.0 × 10 6 cells from each population were inoculated into flasks containing 8.0 ml of medium containing less than 10 nM Fe and 0, 2, 6, 18, or 54/zM adriamycin. As the inocula represented dilution factors of 1:1 to 1:100, at most 1.8 nM Fe was added to the basal medium. The flasks were incubated 26 h; then aliquots of the suspended cells were diluted 1:5 in 0.4% potassium chloride and the cells were counted under a microscope using a hemocytometer.

Reaction of BPS and ascorbate with human iron-trans-

ferrin and fetal bovine serum. The kinetics of iron release from human holo-transferrin were measured by following the increase in absorbance of the Fe(lI) (BPS)3 product at 534 nm as a function of time. The measurements were made on a Beckmann DU-70 spectrophotometer. About 0.5 ml of fetal bovine serum was incubated at 37°C for 24 h in the presence or absence of 0.8 mM BPS or 0.8 mM BPS plus 2.5 mM ascorbate. After 24 h, the samples were spectrophotometrically scanned from 250 to 750 nm for the presence of Fe(II)(BPS)3. The samples were also applied to a calibrated Sephadex G-75 chromatographic column that was equilibrated with 20 mM Tris. HCI buffer. Fractions were collected with 20 mM Tris, pH 7.8 as the eluting buffer. The fractions were analyzed for iron by furnace atomic absorption spectrophotometry, and for zinc by flame atomic absorption spectrophotometry. Samples analyzed for iron were diluted ten times with doubledistilled H20 prior to analysis. The blank and the standard solutions for the iron measurements were made in 2 mM Tris. Distribution of BPS or Fe(II)(BPS)3 between octanol and water. Water containing 20 mM BPS and 1-octanol in a ratio of 1 to 30 ml were equilibrated with frequent shaking for 30 min at 37°C in a separatory funnel. After settling, the layers were separated and 20 mM aqueous Fe 2+ was added to each layer and allowed to equilibrate for another 30 min. Then both the water and the octanol layers were analyzed spectrophotometrically for Fe(II)(BPS)3. A similar experiment was done substituting Fe(II)(BPS)3 for BPS. The 1octanol/H20 partition coefficient was then calculated as [agent]l ........ ~/[agent]H2O × 30. RESULTS

Effect o f BPS and ascorbate on the growth and iron content of HL-60 cells HL-60 cells were incubated for 48 h with various concentrations of BPS, ascorbate, or both BPS and ascorbate to examine the growth inhibitory properties of this ligand in comparison with OP. According to Fig. lb, BPS, with or without ascorbate, displayed only a shallow concentration-dependent inhibition of cell survival up to at least 100 #M BPS, whereas OP was highly toxic even at micromolar concentrations (Fig. la). Fe(II)(BPS)3 had no effect on cell proliferation (Fig. lc). Homogenates and supernatants made for analysis of Fe content and distribution were prepared from nearly confluent iron-normal cells and cells exposed for 24 h

Cell iron deficiency and adfiamycin toxicity 100

100,

.J ,,_.1

t.O O

,,_1 ._j

80

ILl o

a

LLI ..,J

co <->-

60-

.._1

O rr I.z O o Ii

O o~

80-

U.l ._I

co

,<

323

60-

..._1

40-

O rr

40

20

O O LL O

20

0

I.z

,

0

,

,

~

,

,

,

,

,

2 4 6 8 1,10-PHENANTHROLINE,~M

6o ,,_1 ,,_1 LLI

o ILl ,--I

co ,< _,..I

O n" t-. z O O ii

O o~

0

0

5'o

lOO

15o

200

BPS, ~M

100,

90807060504030201000

5'0

160,

150

Fe(BPS)3, }JM

Fig. 1. Effect of BPS, Fe(II)(BPS)3, OP, and ascorbate on proliferation and viability of HL-60 cells. (a) OP, (b) BPS without (11) and with (D) ascorbate, (c) Fe(II)BPS3.

tO BPS and ascorbate. A c c o r d i n g to T a b l e 1, total cellular F e was d e c r e a s e d only when both BPS and ascorbate were present in the incubations. In that case, exposure to 80/.tM BPS and 250 # M ascorbate r e d u c e d the Fe content o f the H L - 6 0 cell population b y 56% on a per-cell basis. H a d the cells s i m p l y not obtained any extracellular Fe o v e r the 24 h period, the total a m o u n t o f F e w o u l d have been u n c h a n g e d and the cellular iron

Table 1. Metal Distribution in HL-60 and H9c2 (2-1) Cells" Iron

HL-60 control HL-60 + 80 /zM BPS HL-60 + 80 #M BPS + 250 #M ASC H9c2(2-1 ) control H9c2 (2-1) + 80 #M BPS + 250/zM ASC a nmol/108 cells.

Zinc

Cytosolic

Total

Cytosolic

Total

8.0

9.1

21

27

7.8

9.2

21

21

2.6 22

4.1 24

24 43

28 77

6.6

12

47

78

concentration d i m i n i s h e d b y 7%, b e c a u s e the stable cell population increased 7% through cell division. Thus, iron was lost from cells in the presence o f BPS and ascorbate. The m a j o r pool o f cytosolic iron, which c o m p r i s e d 88% o f cell b o u n d Fe, was c h r o m a t o g r a p h e d o v e r S e p h a d e x G-75 as a high m o l e c u l a r weight species, and was p r o b a b l y ferritin (Fig. 2a). O n l y in H L - 6 0 cells treated with both BPS and ascorbate did this band almost disappear. BPS alone had no impact on its size. Notably, there was no impact on Zn content or distribution in any o f these incubations (Table 1 and Fig. 2b). C h r o m a t o g r a p h i c e x a m i n a t i o n o f the iron distribution in n o r m a l H9c2 (2-1) cells and cells grown in the presence o f BPS and ascorbate also showed that 70% o f the cytosolic iron was lost in these cells (Fig. 2c), but that cytosolic Zn was u n c h a n g e d (Fig. 2d). Neither visual inspection o f cell pellets nor spectrop h o t o m e t r i c e x a m i n a t i o n o f cell suspensions was able to detect the p i n k i s h - r e d Fe(II)(BPS)3 c o m p l e x in cells p r e v i o u s l y w a s h e d with PBS. Also, no measurable Fe c o m p l e x f o r m e d in h o m o g e n a t e s from cells e x p o s e d to 200 # M BPS for 60 min, w a s h e d free o f extracellular

S. NYAYAPATIet al.

324

0.6

0.40

0.5

a

0.30 _J

(5 =- 0.20

(5 ..)-

Z

Z

ot

b

-~ 0.4-

erritin ._

0.3 0.2-

0.10

0.1J

0 00

~m

10 20 30 FRACTION NUMBER

40

1'o 2'0 30 FRACTION NUMBER

06

0.25

40

0.4 0.35

0.2

(5

0.25

0.15-

(5 1

z" O

d t-

0.1-

¢r

0.2 0.15 0.1

0.05 0

d

0.3

0

5

10 15 20 25 30 35 40 FRACTION NUMBER

0"051 0 0

5

10 15 20 25 30 35 40 FRACTION NUMBER

Fig. 2. Sephadex G-75 chromatographicseparation of cytosol from HL-60 and H9c2 (2-1) cells in the absence (11) and presence (D) of 24-h treatment with 80 #M BPS and 250 #M ascorbate. HL-60metal distribution: (a) iron distribution, (b) zinc distribution. H9c2 (2-1) metal distribution: (c) iron distribution, (d) zinc distribution.

BPS, and then sonicated and reacted with 70/.tM Fe plus 200 #M ascorbate. Thus, there was no evidence of intracellular ligands or Fe(II)(BPS)3 in cells exposed to BPS and ascorbate. Therefore, it was concluded that negligible concentrations of complex equilibrate across the plasma membrane; consequently, Fe(II)(BPS)3 could not be the form of the iron that was released from cells.

Partition coefficients of BPS and Fe(II)(BPS)3 The partition coefficient of Fe(II)(BPS)3 between octanol/H20 is 3.7 × 10 -3 and that of BPS alone is 4.6 × 10 -3. These small values suggest that very little Fe(II)(BPS)3 or BPS can enter or exit cells by passive diffusion through the cell membrane. Overnight equilibration of Fe(II)(BPS)3 or BPS with water and octanol showed no change in partition coefficient values.

Antioxidant concentrations in cells treated with BPS and ascorbate In order to use cells depleted of Fe by BPS and ascorbate for studies of the involvement of iron in oxi-

dant damage, it was important to know whether the treatment affected the concentrations of the components of the cell's antioxidant system. Thus, the concentrations of superoxide dismutase, catalase, glutathione peroxidase, glutathione, and ascorbate were measured in cells made deficient in Fe by incubation with BPS and ascorbate. Table 2 summarizes the results. Total superoxide dismutase, catalase, and glutathione levels were unaltered in iron-deficient cells, while glutathione peroxidase was marginally increased by 9%. A small concentration of ascorbic acid was detected in cells grown in the absence of added ascorbic acid. This concentration was elevated three to four times in the presence of 2 5 0 - 4 0 0 #M added ascorbate. Interestingly, no elevation occurred when the cells were exposed to the combination of BPS and ascorbate (Table 3).

Chelation of iron in fetal calf serum by BPS and ascorbate It was shown previously that the uncharged molecular OP is readily accumulated by cells, is highly toxic

Cell iron deficiency and adriamycin toxicity Table 2. Antioxidant Activities in HL-60 Cells Treated With BPS and Ascorbate BPS/ Ascorbate-Treateda

Antioxidant Total SOD Catalase Glutathione peroxidase GSH ~

3.7 1.2 5.1 1.6

± 1.6 _+ 0.1 _+ 0.1 ± 0.1

band of Zn (Fig. 3b). The production of Fe(II)(BPS)3 was confirmed by spectrophotometric inspection of the product mixture, which revealed a spectral peak at 534 nm characteristic of this complex (not shown).

Control a

(3) (3) (3) h (3)

2.8 1.2 4.6 1.5

_+ 0.8 _+ 0.2 _+ 0.3 ± 0.1

(3) (3) (4) b (3)

Reaction of Fe-transferrin with BPS and ascorbate Ligand substitution reactions of Fe(III)-Tr with competing ligands have commonly involved chelating agents that can interact with the Fe-bound carboxylate or other anion sites on the protein to labilize the Fe(III) sites. 2z23 There have also been reports of ligand substitution reactions coupled to oxidation-reduction processes that reduce Fe(III) to Fe(II), which is bound much less strongly to the transferrin protein. 24'25 It can be seen in Fig. 4 that Fe bound to Tr was readily reduced and transferred to Fe(II)(BPS)3 when Tr-Fe was reacted with BPS and ascorbate. Under these pseudo first-order conditions, product formation occurred in two rate steps, each comprising about 50% of the total reaction. These are thought to represent independent reactions of the two iron-containing domains with BPS and ascorbate. A detailed investigation of such reactions will be described elsewhere (Nyayapati and Petering, unpublished information).

" Cells were treated with 144 #M BPS and 400 #M ascorbate

for 24 h. Units are total enzyme units per 1 0 6 cells for each form of SOD and for glutathione peroxidase, nmol H202 consumed per minute per l06 cells for catalase, and nmol/106 cells for GSH. Numbers in parentheses are numbers of experiments. h Statistically different results according to Student's t-test (p <

0.01). c Oxidized glutathione (GSSG) was less than 1% of GSH in all experiments.

to cells, effectively decreases intracellular Fe content, and also reduces cellular Z n . 3J6 On the premise that BPS may behave differently because its - 2 charge should limit its uptake by cells, the capacity of BPS to remove Fe extracellularly from Fe-Tr in fetal calf serum was examined. As seen in Fig. 3a, Sephadex G75 chromatography of untreated serum revealed the presence of a single band of high molecular weight Fe, thought to represent Fe-Tr. Incubation for 24 h with no treatment or BPS alone did not reduce or shift this pool of Fe into a low molecular weight fraction, which could indicate formation of Fe(II)(BPS)3 (data not shown). However, with the addition of ascorbate to the reaction mixture, most of the high molecular weight pool disappeared with the presumed formation of Fe(II)(BPS)3 (Fig. 3a). The product complex binds to Sephadex, as shown by its elution, more than 20 fractions later than Na ÷, the indicator for the excluded volume of the gel filtration colunm. Under these conditions there was no effect on the high molecular weight

Effect of iron deficiency in the action of adriamycin in three cell types In order to study the role of iron in cellular toxicity caused by adriamycin, four different models of irondeficient cells were examined. HL-60 cells adapted to grow in serum-free medium were made iron deficient by removal of their iron source, Fe-PIH. Iron-deficient cells were also obtained by incubating cells in medium supplemented with 10% serum, BPS, and ascorbate. 0.6

0.35-

~

0.30-

T~

0.25-

t/

020-

I ~

_ 015-

a

0.5

Fe-Tr ::L 0.3z

\

\

o oo ~"~'~'T-'~ ()- . . . . . 1'0

b

0.4-

l '~/

005 "

325

0.20.1

, -

20

3'0-

FRACTION NUMBER

--

40

5

10 15 20 25 30 35 40

FRACTION NUMBER

Fig. 3. Sephadex G-75 chromatography of fetal bovine serum in the absence and presence of 24-h incubation with BPS and ascorbate. (a) Serum iron without (11) and with (El) 800/~M BPS and 2.5 mM ascorbate. (b) Serum zinc with the treatments as in a.

S. NYAYAPATIet al.

326 Table 3. Ascorbate Levels in HL-60 Cells" Treatment Controls Ascorbate Ascorbate Ascorbate Ascorbate

nmol/lO" cells only (250 (250/zM) only (400 (400/zM)

#M) + BPS (80 #M) #M) + BPS (144 #M)

0.2 0.8 0.2 0.9 0.4

+_ 0.2 _+ 0.4 _+ 0.2 _+ 0.6 _+ 0.3

(6)b (4)' (4) (3)'t (2)

" Conditions of exposure were the same as in Table 2. b Numbers in parentheses are number of experiments. Statistically different from controls according to Student's t-test (p < 0.05). '~Statistically different from controls according to Student's t-test (p < 0.1).

Both H L - 6 0 and H9c2 (2-1) cells were treated in this manner. H L - 6 0 cells were also grown in m e d i u m plus the nondepleted, high m o l e c u l a r weight extract o f fetal b o v i n e serum. Finally, Euglena gracilis cells were m a d e iron deficient to various extents by placing them in an iron-depleted m e d i u m s u p p l e m e n t e d with several concentrations o f iron. Each population o f iron-deficient cells was then treated with various concentrations o f adriamycin. In each case, the iron-deficient cells were much less sensitive to a d r i a m y c i n than the iron-normal cells (Fig. 5 a - e ) . Specifically, three different methods to m a k e iron-deficient H L - 6 0 cells reduced the concentrationd e p e n d e n t cytotoxicity o f a d r i a m y c i n (Fig. 5 a - c ) . W h e n iron in the form o f F e - T r was added to the irondeficient cells in Fig. 5c and treated with adriamycin, the cells exhibited similar toxicity to that o f the ironnormal control cells. Interestingly, both iron-normal and iron-deficient cells g r o w n in serum-containing med i u m were substantially more sensitive to the drug than either type o f cell adapted to the serum-free m e d i u m . The large disparity in activity o f a d r i a m y c i n in irondeficient and iron-normal cells was also clearly seen in rat heart m y o b l a s t s (Fig. 5d) and in the Euglena gracilis cells (Fig. 5e). Indeed, with the Euglena gracilis cells it was possible to show a sharp d e p e n d e n c e o f cytotoxicity on the iron concentration to which the cells had been adapted. No inhibition occurred at 0 or 9 n M a d d e d Fe, whereas 18 n M F e supported almost the full cytotoxicity o f the drug. This effect was not due to differences in extracellular Fe, because all o f the cultures were transferred to a c o m m o n m e d i u m containing 10 n M Fe prior to addition o f adriamycin.

intracellular iron. F o r example, studies o f the control o f expression o f transferrin receptor and ferritin proteins by iron status have used d e s f e r r i o x a m i n e to m o d ulate intracellular iron. 26'27 Investigations o f the role o f intracellular iron in oxidant stress have utilized OP and D F O to inhibit cytotoxicity caused by hydrogen p e r o x i d e and agents that m a y generate H202 in cells. 12~'2'~ Finally, to e x a m i n e whether the antitumor agent b l e o m y c i n requires intracellular F e for its cytotoxic properties, it has been shown that 1,10-phenanthroline reduces D N A d a m a g e caused b y the drug. 3° The use o f chelating agents to lower intracellular iron content introduces a m b i g u i t y into the interpretation o f the subsequent effects o f iron depletion, because once they enter cells, they might interact in multiple w a y s with cellular components. T h e y might bind either to other metal ions, causing cellular responses that are not related to the presence of Fe, or react with cellular constituents in w a y s unrelated to their capacity to bind metal ions, for e x a m p l e , O P can bind to D N A . 3~ Thus, m o d e l s have been d e v e l o p e d that only l o w e r extracellular nutrient Fe, such as the Euglena gracilis and H L 60 cell systems described in the Introduction. ~'1~ H o w ever, the approaches used with these m o d e l s are limited to cells that can be grown in rigorously defined growth m e d i a in the absence o f serum. The experiments reported here characterize the effects o f BPS and ascorbate on H L - 6 0 cells, which show whether or not these reagents p r o v i d e an effective alternative to 1,10-phenanthroline or d e s f e r r i o x a m i n e for the control o f intracellular iron in cells that must be g r o w n in m e d i a containing serum. One o f the promis-

I.~)

There are numerous current research p r o b l e m s that focus on iron m e t a b o l i s m and distribution in cells and in which metal-chelating agents are e m p l o y e d to alter

I

I

]

I

1.2 I.O 1.8 o 0.6 '~ 0.4 0.2 0 350

DISCUSSION

I

1.4

400

450 500 550 WAVELENGTH, NM

600

650

Fig. 4. Absorbance changes during the reaction of iron-transferrin with BPS and ascorbate. The reaction mixture was 0.05 mM in diferric transferrin, 4 mM in BPS, and 50 mM in ascorbate in 50 mM Hepes, pH 7.4 at 20°C.

Cell iron deficiency and adriamycin toxicity .3-

1 oo

0 n," (.9

80

0r~r

60

_.1 u.I

,,_1 .._1 u,.I

(.)

"1-

--

(..)

,,_1

O

n,i,z

0 o

k l-

20

a

0.0

T I--

O n-(.5 ,,_1 .._1 I.U

o

80

0 0.0

•"r

100,

orr

8O

0'.1 012 0'.3 0'.4 ADRIAMYCIN, ~M

0.5

....I

tu O

60

60-

.._1

20

O rr Iz O o

0

O o~

40

Z

o

20-

100

,,_1

ot r

40

1.0

b

80 60

o

0'.2 0.4 016 0'.8 ADRIAMYCIN, .pM

1O0

F.Z

o

0 o~

120

-

o

40

327

i1

i,

0

0

0.1 0.2 0.3 0.4 ADRIAMYCIN, .pM

T

140

O rr"

120-

J J LU

o

0.5

4020-

d 0 0.0

0.2 0.4 ADRIAMYCIN, #M

0.6

10080-

.d

O rr t-Z

O (..) u_ O o~

6040200 15

30

45

60

ADRIAMYCIN, ~IM

Fig. 5. Effects of adriamycin on proliferation and viability of three cell types. (a) HL-60 cells, control (m) and cells made irondeficient by removal of iron pyridoxal isonicotinoyl hydrazone from the growth medium for 48 h (D). (b) HL-60, control cells (m), and cells treated with 80/~M BPS and 250/~M ascorbate for 24 h (D). (c) HL-60, control cells (m) and cells treated with growth medium and iron-depleted, high molecular weight fraction of fetal bovine serum (E]) for 24 h. (d) H9c2 (2-1), control cells (m), and cells incubated with 80 #M BPS and 250 #M ascorbate for 24 h (E3). (e) Euglena gracilis cells, control cells (m), and cells grown in 9 nM (&), 18 nM (X), 36 nM ([]) and 180 nM (*) iron prior to exposure to Adr.

ing features o f this t r e a t m e n t is that cells lose total F e r a p i d l y after e x p o s u r e to B P S and a s c o r b a t e ( T a b l e 1). S o m e d e c r e a s e in the rate o f cell p r o l i f e r a t i o n o c c u r s as s h o w n in Fig. 1, w h i c h m a y b e r e l a t e d to the loss o f c e l l u l a r iron. A c c u m u l a t i o n o f the n e g a t i v e l y

c h a r g e d B P S m o l e c u l e b y cells or f o r m a t i o n o f intrac e l l u l a r Fe(II)(BPS)3 w a s not o b s e r v e d , s u p p o r t i n g the v i e w that B P S and a s c o r b a t e act in the e x t r a c e l l u l a r m e d i u m . T h i s c o n c l u s i o n w a s s u p p o r t e d by partition c o e f f i c i e n t m e a s u r e m e n t s that d e m o n s t r a t e d as e x -

328

S. NYAYAPATIet al.

pected that BPS 2- and Fe(II)(BPS)3 4 strongly prefer an aqueous solution over a nonpolar medium, mimicking the interior of the cell membrane. Thus, BPS may not directly affect any intracellular process except as a secondary result of a reduction of extracellular iron. The findings that superoxide dismutase, catalase, and glutathione levels in HL-60 cells were hardly altered by exposure to BPS and ascorbate are important to the use of this model in studies of the involvement of intracellular Fe in oxidant damage (Table 2). Their stability shows that any observed effects of iron deficiency on oxidant activity are not due to modification of the concentration of antioxidants in the cells. The extracellular action of BPS and ascorbate is also consistent with the findings that incubation of fetal bovine serum with this combination of agents, but not with BPS alone, caused the loss of high molecular weight Fe presumed to be Fe-transferrin (Fig. 3). Furthermore, purified Fe-transferrin readily reacts with BPS or OP in the presence of ascorbate (Fig. 4). OP reacted directly in cells, whereas BPS caused its effects on intracellular iron through reductive removal of Fe from extracellular transferrin, which is supported by the requirement of ascorbate as an extracellular reductant for BPS. OP lowered cellular Fe without addition of a reductant.~6 Finally, cells exposed to BPS plus ascorbate remained proliferative even though they lost much of their iron stores (Figs. 1 and 2). Not only was ferritin Fe mobilized over the time course of the experiment, it was also readily exported from the cells as seen by the rapid loss of total Fe from the entire cell population. Because Fe(II)(BPS)3 was not accumulated by cells, and similarly would not diffuse out of cells, the mechanism of release of Fe probably did not involve the formation of intracellular Fe(II)(BPS)3. It is important to note that use of BPS and ascorbate to probe the role of iron in cellular damage by activated oxygen or by redox-cycling compounds must still be performed with some degree of caution. Ascorbic acid is an important antioxidant, which can counteract some of the damaging action of free radicals. 32 However, it might also react with iron mobilized by BPS and 02 to generate reactive oxy-radical species. Furthermore, the presence of BPS in the system could perturb cells in ways not yet detected. Thus, the finding that HL60 cells proliferated well in medium plus the macromolecular fraction of serum permitted another means to make cells iron deficient, which overcame the above criticisms. Serum was first reacted with BPS plus ascorbate; then its complement of macromolecules was separated from BPS, Fe(II)(BPS)3, and ascorbate by gel filtration. Finally, this fraction of iron-depleted macromolecules was used as the serum component of

the growth medium. Cells lost iron and displayed reduced sensitivity to adriamycin in the absence of BPS or ascorbate. The fact that addition of Fe-transferrin to this system restored sensitivity to Adr demonstrates conclusively that the key target of BPS + ascorbate is the iron in transferrin. It was previously shown that the limitation of iron in defined growth media leads to iron deficiency in HL-60 and Euglena gracilis cells. 8'16In turn, such cells show a reduced sensitivity to H202 and bleomycin. With H2Oz, one hypothesizes that reduction in cellular iron has decreased a pool of iron that acts as a Fenton catalyst for the reduction of H202 to hydroxyl radical and hydroxide ion. 8 With bleomycin, it is thought that the drug is directly deprived of chelatable iron needed to catalyze the reductive activation of O~ that causes DNA strand scission. 8"33 Using these same two models plus the new ones, which permit the examination of iron deficiency in cells grown in serum-containing medium, it is now evident that the cytotoxicity of adriamycin, too, is highly dependent on the presence of cellular iron (Fig. 5). Indeed, in Euglena gracilis its action seems to require cellular iron. Adriamycin has substantial affinity for Fe(III) and is also redox active/"~ ~.~2 Thus, it has been attractive to hypothesize that adriamycin utilizes iron directly or indirectly to generate activated forms of oxygen such as hydroxyl radical, which cause cytotoxicity, u The present results provide strong support for this view. The effects of adriamycin are not completely abolished in iron-deficient, mammalian cells because residual iron is still present in these cells to sustain critical metabolic systems. Experiments also point to the inhibition of topoisomerase II by adriamycin as a mechanism of cell injury for this drug. 15'34The present results do not disprove this hypothesis, at least in mammalian cells in which the condition of iron deficiency does not completely eliminate the activity of the drug. Possibly, the residual toxicity is either due to inhibition of topoisomerase II, or a reaction with topisomerase II, which might require iron in some unknown way. Nevertheless, the connection between cellular iron and much of the cytotoxicity of adriamycin in three diverse cell types has been established. - - T h e authors acknowledge support from an American Cancer Society Grant ACS-DHPd, NIH Grant R29CA58611 (R.W.B.), and the University of Wisconsin-Milwaukee Graduate School (G.A.).

Acknowledgements

REFERENCES

1. de Mello Felho, A. C.; Meneghini, R. Protection of mammalian cells by o-phenanthroline from lethal and DNA-damaging ef-

Cell iron deficiency and adriamycin toxicity

2. 3.

4.

5. 6.

7.

8.

9.

10. 1 I. 12.

13. 14.

15. 16.

17.

18. 19.

fects produced by active oxygen species. Biochem. Biophys. Acta 847:82-89; 1985 lmlay, J. A.; Chin, S. M.; Linn, S. Toxic DNA damage by hydrogen peroxide through the Fenton reaction in vivo and in vitro. Science 240:640-642; 1988. Krishnamurti, C.; Saryan, L. A.; Petering, D. H. Effects of ethylenediaminetetraacetic acid and 1,10-phenanthroline on cell proliferation and DNA synthesis of Ehrlich ascites cells. Cancer Res. 40:4092-4099; 1980. Richardson, D.; Ponka, P.; Baker, E. The effect of the iron(Ill) chelator, desferrioxamine, on iron and transferrin uptake by the human malignant melanoma cell. Cancer Res. 54:685-689; 1994. Sillen, L. G.; Martell, A. E. Stability constants supplement No. 1. London: The Chemical Society Burlington House; 1971:676677. Propper, R. D.; Shurin, S. B.; Nathan, D. G. Desferrioxamine B and iron overload. In: Anderson, W. F.; Hiller, M. C., eds. Development of iron chelators for clinical use. Washington, DC: DHEW Publication No. (NIH) 76-994; 1975:83-114. Byers, B. R.; Arceneauz, J. E. L.; Gaines, B. G.; Sciortino, C. V. Isolation of microbial iron chelators: Some possible effects of their chemotherapeutic use. In: Anderson, W. F.; Hiller, M. C., eds. Development of iron chelators for clinical use. Washington, DC: DHEW Publication No. (NIH) 76-994; 1975:213-228. Radtke, K.; Lomitzo, F. A.; Byrnes, R. W.; Antholine, W. E.; Petering, D. H. Iron requirement for cellular DNA damage and growth inhibition by hydrogen peroxide and bleomycin. Biochem. J. 302:655-664; 1994. Keizer, H. G.; Pinedo, H. M.; Schurhuis, G. J.; Joenje, H. Doxorubicin (Adriamycin): A critical review of free radical-dependent mechanisms of cytotoxicity. Pharmacol. Ther. 47:219231; 1990. Wallace, K. B. Nonenzymatic oxygen activation and stimulation of lipid peroxidation by doxorubicin-copper. Toxicol. Appl. Pharmacol. 86:69-79; 1986. Gelvan, D.; Samuni, A. Reappraisal of the association between adriamycin and iron. Cancer Res. 48:5645-5649; 1988. Massoud, S. S.; Jordan, R. B. Kinetic and equilibrium studies of the complexation of aqueous iron(III) by daunomycin, quinizarin, and quinizarin-2-sulfonate, lnorgan. Chem. 30:48514856; 1991. Eliot, H.; Gianni, L.; Myers, C. Oxidative destruction of DNA by the adriamycin-iron complex. Biochemistry 23:928-936; 1984. Gianni, L.; Zweier, L. L.; Levy, A.; Myers C. E. Characterization of the cycle of iron-mediated electron transfer from adriamycin to molecular oxygen. J. Biol. Chem. 260:6820-6828; 1985. Tewey, K. M.; Rowe, T. C.; Yang, L.; Halligan, B. D.; Liu, L. F. Adriamycin-induced DNA damage mediated by mammalian DNA topoisomerase I1. Science 226:466-468; 1984. Lyman, S.; Taylor, P.; Lornitzo, F.; Weir, A.; Stone, D.; Antholine, W. E.; Petering, D. H. Activity of bleomycin in iron- and copper-deficient cells. Biochem. Pharmacol. 38:4273-4282; 1989. Loven, P.; Leeper, D. B.; Oberley, L. W. Superoxide dismutase levels in Chinese hamster ovary cells and ovarian carcinoma cells after hyperthermia or exposure to cycloheximide. Cancer Res. 45:3029-3033; 1985. Aebi, H. E. Catalase. In: Bergmeyer, H. U., ed. Methods of enzymatic analysis, 3rd ed. vol. III. Weinheim: Verlag-Chemie; 1983:277-282. Geiger, P. G.; Thomas, J. P.; Girotti, A. W. Lethal damage to murine L1210 cells by exogenous lipid hydroperoxides: Protective role of glutathione-dependent selenoperoxidases. Arch. Biochem. Biophys. 288:671-680; 1991.

329

20. Griffith, O. W. Glutathione and glutathione disulphide. In: Bergmeyer, H. U., ed. Methods of enzymatic analysis, 3rd ed. vol. 9. Weinheim: Verlag-Chemie; 1985:521-529. 21. Omaye, D. W.; Emery, G.; Maynard, J. E. Selected methods for the determination of ascorbic acid in animal cells, tissues, and fluids. Methods Enzymol. 62:3-11; 1979. 22. Marques, H. M.; Watson, D. 1.; Egan, T. J. Kinetics of iron removal from human serum monoferric transferrins by citrate. lnorg. Chem. 30:3756-3762; 1991. 23. Egan, T.; Ross, D. C.; Purves, L. R.; Adams, P. A. Mechanism of iron release from human serum C-terminal monoferric transferrin to pyrophosphate: Kinetic discrimination between alternative mechanisms, lnorg. Chem. 31:1994-1998; 1992. 24. Kojima; Bates, G. The reduction and release of iron from Fe 3+transferrin-CO32-. J. Biol. Chem. 254:8847-8854; 1979. 25. Ankel, E.; Petering, D. H. Iron-chelating agents and the reductive removal of iron from transferrin. Biochem. PharmacoL 29:1833-1837; 1980. 26. Chitambar, C. R.; Zivkovic-Gilgenbach, Z. Influence of cellular iron status on the release of soluble transferrin receptor from human promyelocytic leukemic HL-60 cells. J. Lab. Clin. Med. 116:345-353; 1990. 27. Rodgers, J. T.; Bridges, K. R.; Durmowicz, G. P.; Glass, J.; Auron, P. E.; Munro, H. N. Translational control during the acute phase response. Ferritin synthesis in response to interleukin-l. J. Biol Chem. 265:14572-14578; 1990. 28. de Mello Filho, A. C.; Hoffmann, M. E.; Meneghini, R. Cell killing and DNA damage by hydrogen peroxide are mediated by intracellular iron. Biochem. J. 218:273-275; 1984. 29. Shiloh, H.; lancu, T. C.; Bauminger, E. R.; Link, G.; Pinson, A.; Hershko, C. Desferoxamine-induced iron mobilization and redistribution of myocardial iron in cultured rat heart cells: Studies of the chelatable iron pool by electron microscopy and Mrssbauer spectroscopy. J. Lab. Clin. Med. 114:428-436; 1992. 30. Byrnes, R. W.; Petering, D. H. Inhibition of bleomycin-induced cellular DNA strand scission by 1,10,phenanthroline. Biochem. Pharmacol. 41:1241-1248; 1991. 31. Graham, D. R.; Sigman, D. S. Zinc ion in Escherichia coli DNA polymerase: A reinvestigation, lnorg. Chem. 23:4188-4191; 1984. 32. Winkler, B. S.; Orselli, S. M.; Rex, T. S. The redox couple between glutathione and ascorbic acid: A chemical and physiological perspective. Fr. Radiat. Biol. Med. 17:333-349: 1994. 33. Petering, D. H.; Byrnes, R. W.; Antholine, W. E. The role of redox-active metals in the mechanism of action of bleomycin. Chem.-Biol. Interact. 73: ! 33 - 182; 1990. 34. Defile, A. M.; Batra, J. K.; Goldenberg, G. J. Direct correlation between DNA topoisomerase II activity and cytotoxicity in adriamycin-sensitive and -resistant P388 leukemia cell lines. Cancer Res. 49:58-62: 1989.

ABBREVIATIONS

Adr--adriamycin BPS--bathophenanthrolinedisulfonate or 4,7-phenylsulfonyl- 1,10-phenanthroline Fe-PIH--iron-pyridoxal isonicotinoyl hydrazone GSH--glutathione GSSG--oxidized gluthatione NBT--nitro blue tetrazolium OP-- 1,10-phenanthroline

PBS-- phosphate-buffered saline SOD--superoxide dismutase Tr--transferrin