[CANCER RESEARCH 45, 5257-5262, November 1985]

Modulation of Cytotoxicity of Menadione Sodium Bisulfite versus Leukemia L1210 by the Acid-soluble Thiol Pool1

StevenA.Akman,2MarianDietrich,RowanChlebowski,PhyllisLimberg,andJeromeB.Block

Division of Medical Oncology, UCLA School of Medicine, Harbor-UCLA Medical Center, Terranee, California 90509

ABSTRACT growth inhibition by menadione is not known. Due to its quinone structure, menadione participates as a powerful oxidation-reduc We investigated the mechanism of antitumor activity of the tion agent in oxygen-consuming electron transfer reactions cat water-soluble derivative of menadione, menadione sodium bisul alyzed by microsomal enzymes (8-12). Menadione semiquinone fite (vitamin KJ, versus murine leukemia L1210. Vitamin Ka, in free radical is generated during reduction of the quinone (9,10); concentrations >27 ¿¿M,causedtime- and concentration-depend this radical is short lived, producing Superoxide radicals (9, 10, ent depletion of the acid-soluble thiol (GSH) pool. Maximal GSH 12, 13). The consequences of menadione reduction have been depletion to 15% of control occurred at 45 /¿Mvitamin«3.Vitamin invoked as the cause(s) of menadione toxicity in normal hepa- Ka-mediated GSH depletion and vitamin Ks-mediated growth tocytes (8, 13) and neutrophils (14), and menadione is thought inhibition were abrogated by coincubation with 1 mw cysteine or to be responsible for: (a) depletion of NAD(P)H with subsequent 1 mw reduced glutathione but not by 1 mw ascorbic acid or 180 depletion of mitochondrial ATP (15); (b) Superoxide and peroxy fiM a-tocopherol. Low concentrations of vitamin Ka (9-27 ¿¿M)radical generation, causing peroxidative damage (13, 16, 17); elevated both the GSH pool and the total glutathione pool, the and (c) depletion of cellular GSH3 (13, 14). The data presented latter to a greater degree. Vitamin Ka also caused an increased in this paper suggest that menadione potently depletes tumor rate of Superoxide anióngeneration by L1210, maximal at 45 I*M cell GSH and NADP pools and that acid-soluble thiols are impor vitamin Ka (300% of control), and a concentration-dependent tant modulators of menadione antitumor activity. depletion of the reduced nicotinamide adenine dinucleotide phos phate (NADPH) and total nicotinamide adenine dinucleotide phos phate (NADP) pools. Forty-fifty % depletion of the NADPH pool MATERIALS AND METHODS occurred after exposure to 27 /¿Mvitamin«3and 100% occurred Chemicals and Reagents. L-Cysteine, vitamin K3, ascorbic acid, at 36 nu vitamin Ka; 27 /tw vitamin «3isa nontoxic concentration NADPH, NADP+, reduced glutathione, Superoxide dismutase (EC of vitamin Ka. Loss of NADPH and total NADP was prevented 1.15.1.1, from bovine blood), glutathione reductase (EC 1.6.4.2), and all by coincubation with 1 mw cysteine but not by coincubation with reagents for the assay for pyridine nucleotides were purchased from ascorbic acid or a-tocopherol. We conclude that tumor cell Sigma Chemical Co. (St. Louis, MO). a-Tocopherol:polyethylene glycol growth inhibition by vitamin Kais modulated by acid-soluble thiols 100:succinate was purchased from Eastman Chemical Co. (Kingsport, and may be caused by GSH pool and/or NADPH depletion. TN). This derivative of a-tocopherol is soluble in boiling water at concen Toleration of partial NADPH depletion by L1210 cells may indi trations «50 mW. DTNB was purchased from K and K Laboratories cate that a threshold level of NADPH loss of >50% is necessary (Hollywood, CA). Lapachol [1,4-naphthoquinone-2-hydroxy-3-(methyl- for toxicity. NADPH depletion may be a toxic effect common to 2-butenyl), NSC 11905] and dichlorolapachol [2-hydroxy-3-(3,3-dichlo- quinone drugs. Equitoxic concentrations of vitamin K;1, phyllo- roallylH ,4-naphthoquinone, NSC 125771] were received as gifts from the Drug Synthesis and Chemistry Branch, Division of Cancer Treatment, quinone, lapacho!, dichlorolapachol, and doxorubicin caused National Cancer Institute, Bethesda, MD. Doxorubicin was obtained from L1210 NADPH pools to deplete to 30 ±10 (SD), 60 ±10, 60 ± Adria Laboratories (Columbus, OH). Phylloquinone (Aqua Mephyton) was 11, and 80 ± 12% of control, respectively. In contrast, GSH obtained from Merck, Sharpe, and Dohme (Westpoint, PA). All chemicals depletion may not be a common mechanism of toxicity. Of these were used as received. quiñones,only vitamin K3caused significant GSH depletion when Cell Culture. Leukemia L1210, a murine leukemia maintained in long studied in equitoxic concentrations. term liquid suspension culture, was used in all experiments. L1210 cells were maintained in 15-ml tissue culture flasks (Falcon Plastics, Oxnard, CA) at an initial concentration of 100,000 cells/ml using Roswell Park INTRODUCTION Memorial Institute Medium 1640 (Grand Island Biological Co., Grand Island, NY) supplemented with 15% heat-inactivated (56°Cfor 30 min) The naphthoquinone derivative menadione (2-methyl-1,4- fetal bovine serum, penicillin (100 units/ml), streptomycin (100 ng/m\), naphthoquinone) has been shown to inhibit the growth of tumor and i -glutamina (2 mw) at 37°C in an atmosphere containing 5% CO2 cells in vitro (1-6). In the human tumor stem cell soft agar cloning and 95% air. Cultures were fed twice weekly and were in logarithmic assay (7), menadione causes inhibition of clonal growth of a wide growth phase at the time of all experimental studies. variety of tumor cell types (3). Its anticancer activity with this Growth inhibition by the drugs reported in these studies were observed by the dose-response method described previously (18). Briefly experi assay is comparable or superior to currently used standard ments were carried out in 15-ml tissue culture flasks. Supplemented chemotherapeutic agents, such as doxorubicin. Menadione is in Roswell Park Memorial Institute Medium 1640, drug(s), and cells in use in early trials in patients with advanced cancer (2) with the logarithmic growth were added to the flasks and incubated at 37°C in antimetabolite 5-fluorouracil. The mechanism of tumor cell 'The abbreviations used are: GSH, total-acid soluble thiols, reduced form; 1Supported in part by a grant from the California Institute For Cancer Research. vitamin K3. menadione sodium bisulfite; GSSG, total acid soluble thiols, oxidized 2To whom requests for reprints should be addressed. form; DTNB, 5,5'-dithiobis-<2-nitrobenzoic acid); SOD, Superoxide dismutase (EC Received 9/4/84; revised 7/24/85; accepted 7/26/85. 1.15.1.1).

CANCER RESEARCH VOL. 45 NOVEMBER 1985 5257

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1985 American Association for Cancer Research. MODULATION OF REDUCED GLUTATHIONE BY VITAMIN

an atmosphere of 5% CO2 with 95% air. Cells were added at an initial centrifugation at 280 x g for 10 min at 4°C. Cells were then washed concentration of 100,000 cells/ml. Drugs were diluted in Dulbecco's three times in ice-cold Dulbecco's phosphate-buffered saline (19) supple phosphate-buffered saline (Grand Island Biological Co.) (19). Each ex mented with 10 mM glucose. After cells were washed, NADPH pools perimental point was derived from triplicate cultures. After 16 h of were determined on cells which were extracted with 300 n\ of tee-cold incubation, cells were washed free of drug with two washes of Ros well 0.5 N KOH in 50% ethanol, heated for 5 min at 90°C, recooled to 0°C, Park Memorial Institute Medium 1640 and incubated in fresh supple and neutralized with 0.5 M triethanolamine in 0.4 M dibaste potassium mented media. Cells were fed with new media on day 4. Cell counts phosphate. NADPH in the neutralized alkaline extracts was converted to were performed on days 4 and 7 on a Model ZBI Coulter Counter NADP* prior to assay by addition of 10 MM GSSG and 0.2 unit of (Coulter Electronics, Hiateah, FL). glutathione reductase (Sigma type III, from yeast) and incubation at 37°C Acid-soluble Thiols. Acid-soluble thiol pools were determined for for 30 min. The extracts were then acidified with ice-cold 2 M m- aliquot s of 10 to 50,000,000 cells by modification of the method of phosphoric acid, centrifuged at 15,000 x g for 5 min to remove protein, Beutter ef al. (20). Cells were cooled on tee for 5 min and harvested by and reneutralized with 2 N KOH in 0.3 M N-morpholinopropanesulfonic centrifugation at 280 x g at 4°C. After removal of aliquots for protein acid. NADP4 pools were determined on cells extracted with 0.2 M m- assay by the method of Bradford (21), the resultant cell pellets were phosphoric acid. After centrifugation at 1500 x g and neutralization of layered onto 600 n\ of 0.14 M m-phosphoric acid containing 4.6 HIM the supematants with 1 N KOH in 0.15 M W-morpholinopropanesulfonic EDTA and 3.47 M sodium chloride, overlaid with 500 ¿ilof silicone fluid acid, NADP* content was determined by measuring the rate of increase DC550 (Dow Chemical, Midland, Ml):light mineral oil, 84:16, in 1.5-ml of absorbance of 570 nm light due to reduction of thiazolyi blue in the Eppendorf centrifuge tubes. Tubes were centrifuged at 15,000 x g for cycling system containing 300 //I of the cell extracts, 0.1 N Tris-HCI (pH 10 s in a microcentrifuge (Fisher Scientific, Tustin, CA), allowing 100% 7.8), 5 mM glucose 6-phosphate, 5 mM MgSO,, 0.8 mM phenazine of the cells to migrate into the m-phosphoric acid layer, with less than methylsulfate, 0.2 mM thiazolyi blue, and 0.3 unit glucose-6-phosphate 2% contamination by supernatant (22, 23). Supernatant and oil layers dehydrogenase (Sigma type IX, from yeast). Each assay was carried out were removed and cell pellets were dispersed in the m-phosphoric acid at 25°C in cuvets containing 1.2 ml total volume. NADP* content was solution by gentle homogenization with a precooled glass rod. Paniculate determined by addition of authentic NADP* standard to each assay. This debris was removed by centntugation at 15,000 x g for 10 min. Aliquots assay was linear in the range 50-2000 pmol. The lower limit of sensitivity (500 /il) of supernatant were mixed with 1.5 ml of a reaction mixture was 10 pmol for acid-extracted samples and 100 pmol for alcoholic containing DTNB (Ellman's reagent; 0.05 mg/ml) in 0.25% sodium citrate alkali-extracted samples. and 0.3 M dibasic sodium phosphate. The absorbance of the reaction mix at 412 nm was determined on a Model 25 double beam spectropho- tometer (Beckman Instruments, Fullerton, CA). RESULTS Total glutathkxie content of 10 to 20 million log phase growth cells were determined by a modification of the method of Akerboom and Sies Incubation with vitamin K3caused a time- and concentration- (24). Cell suspensions were placed in Eppendorf centrifuge tubes con dependent depletion of the GSH pool (Charts 1 and 2). The GSH taining 500 n\ of ice-cold 0.6 N perchloric acid overlaid with 500 ¡Aof pool reached a minimum of 15% of control after 16 h of exposure DC550 silicone fluid:mineral oil (84:16) and centrifuged at 15,000 x g for to 72 fiM vitamin «3.Theconcentration-response relationship for 15 s, after which the oil layers were removed. The acid extracts were GSH depletion was steep, with 16 h exposure to up to 27 *IM neutralized with 1 N KOH in 0.3 M A/-morpholinopropanesulfonic acid vitamin «3producingno depletion and >45 JIMvitamin «3pro buffer. KCIO4 precipitates were removed by centrifugation and 50-^1 ducing maximal depletion of 85%. Low concentrations of vitamin aliquots of the supernatant s were placed in cuvets containing 1 ml of 0.1 Ka(<27 /tM) actually enhanced the GSH pool, by as much as 2- M potassium phosphate buffer (pH 7.0), 0.7 UM NADPH, 20 IM DTNB (Ellman's reagent), and 0.2 unit of glutathione reducíase (Sigma type 3, fokj. from yeast). Loss of NADPH fluorescence in the cuvets (excitation wavelength, 360 nm; emission wavelength, 460 nm) was recorded using an Aminco-Bowman photofluorometer (Travenol, Inc., Elk Park, IL). Glutathione content per cuvet was calibrated to the rate of NADPH lo.o-i disappearance by addition of authentic GSSG standard to each cuvet. Protein content per cell suspension was assayed by the method of Bradford (21 ), using bovine serum albumin type V as standard. « 8. OH Superoxide Generation. Superoxide generation by L1210 cells were 0 assayed by the method of Smith et al. (12). One million cells per assay a. were suspended in 1.35 ml of Krebs-Henseleit buffer plus or minus SOD o> (0.02 mg/ml) and warmed at 37°C for 2 min. Fifty /J of cytochrome c E 6.0- (Sigma type VI, from horse heart; 30 mg/ml) plus or minus 150 fit of various concentrations of vitamin «3were added and the suspensions were incubated for an additional 30 min at 37°C. After incubation, 4. OH suspensions were cooled on tee for 5 min, and 1-ml aliquots were layered over 500 pi of a mixture of DC550 silicone fluid:mineral oil (84:16) in 1.5- ml Eppendorf tubes and centrifuged at 15,000 x g for 10 s to remove 2.0- co 100% of the cells. Absorbance at 550 nm of the supernatant was assayed o immediately. The absorbance due to Superoxide generation was calcu lated as the difference between the absorbance in the absence of SOD H h and the absorbance in the presence of SOD. Concentration of Superoxide was determined using the molar extinction coefficient Am = 21.0 x 10~3 8 I6 M"' cm-1 (25). TIME (hr) NADPH:NADP* Pools. Pyridine nucleotide pools were assayed by a Chart 1. GSH pods were measured by the method of Beutler et al. (20) after incubation with 72 pM vitamin K3 for the times shown. Proteins were assayed by modification of the cycling spectrophotometric assay of Bemofsky and the method of Bradford (21). Points, mean (n = 2); oars, SO. The GSH pool of Swan (26). From 10 to 20 million cells per data point were harvested by control cells was 9.9 ±1.2 ¿/g/mgprotein.

CANCER RESEARCH VOL. 45 NOVEMBER 1985 5258

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1985 American Association for Cancer Research. MODULATION OF REDUCED GLUTATHIONE BY VITAMIN «3

Table 1 Effect of reducing agents on vitamin Ka-mediateddepletion of the LI 210 acid-soluble thtol pool GSH pools were measured after a 16-h exposure to vitamin K3 ±reducing agents by the method of Beutler er al. (20). Proteins were assayed by the method of Bradford (21). Values shown are representative of at least three experiments. The interexperiment coefficient of variation is <12% for the first five and the ninth and tenth data points listed and <36% for the sixth through the eighth data points. Vitamin Ka pool »IM)NoneAscorbicReducingagent (45 Gig/mgprotein)811(1 OOf13(120)13(120)10(90)20(180)4(30)3(20)5(50)23 rriM)a-Tocopherolacid (1.2 -Reduced (0.2 mm) -Cysteineglutathione (1 HIM) (1mu)None +Ascorbic +o-Tocopherolacid 0 3.6 9 18 27 36 45 +Cysteine +Reduced (200)8(70) [KJ] (/IMI glutathione +GSH * Units are calibrated to ¿igofa reduced glutathione standard which produces an equivalentreduction of DTNB. Chart 2. GSH pools (•)weremeasured by the method of Beutler ef al. (20) " Numbers in parentheses,percentageof control. after 16 h incubation with vitamin «3inthe concentrations shown. Total glutathione pools (O) were determined by the methods of Akerboom and Sies (24). Proteins were assayed by the method of Bradford (21). Points, means of the percentage of control (no vitamin «3)values(n = 6); oars, SD. The GSH pool of control cells varied from 12 to 20 n9 of GSH per mg protein. The total glutathione pool varied - IOO- from 16 to 30 nmo! of glutathione per mg protein.

Low concentrations of vitamin «3also elevate the total gluta 80- thione pool (Chart 2). The percentage of elevation of the total glutathione pool at 9-27 n»vitamin K3 was greater than the 60- percentage of elevation of the GSH pod. Also the total glutathi o one pool remained at levels equivalent to those of control cells u after exposure to higher concentrations of vitamin K3 (>36 /IM), 40- despite depletion of the GSH pool. L1210 cells readily repleted the GSH pool after a short term 20- (1 h) incubation with vitamin K3. Within 3 h after removal from a 1-h exposure to media containing 72 ^M vitamin K3, the GSH pool recovered to 90 ± 5% (SD) of control (n = 3) from a 0 3.6 minimum of 60 ±4% of control. However, L1210 cells showed [K.] only a limited capacity to replete the GSH pool after a prolonged (16-h) exposure to vitamin Kg. Five h after removal of cells from Chart 3. Growth curves were generated by placing cells in fresh medium after a 16-h exposure to the appropriate vitamin K3concentrations (see "Materials and a 16-h exposure to medium containing 36 UM vitamin «3,GSH Methods"). Points, mean of the percentageof control (no vitamin K3)cell counts at levels recovered to only 30 ±13% of control from 10 ±2%. No 96 h (n = 2); oars, SD. further recovery of GSH was observed beyond 5 h; the viability of the cell began to deteriorate at that point. the nonthiol reducing agents ascorbic acid and a-tocopherol did Vitamin K3 more readily depleted the L1210 GSH pool than not prevent vitamin K3-mediated GSH depletion (Table 1). did equitoxic concentrations of some other quiñones.Exposure Growth Inhibition. The effect of 16 h exposure to vitamin «3 to 72 MMfor 1 h vitamin «3reduced the GSH pool to 20 ±18% on L1210 growth in liquid suspension culture was evaluated. of control, whereas 53 /*M lapachol (110 ± 9% of control), Exposure for 16 h was chosen to assure maximal depletion of dichlorolapachol (90 ±10% of control), and 37 UM doxorubicin the GSH pool. Vitamin «3inhibited L1210 growth in concentra (110 ±14% of control), all equitoxic at the noted concentrations tions >36 nu (Chart 3). The dose concentration-response rela to L1210 growing in suspension culture (data not shown), did tionship was quite steep with concentrations <36 A/Mhaving no not reduce the GSH pool at all. The equitoxic concentration of effect. phylloquinone (670 MM) actually raised the GSH pool to 150 ± Coincubation of L1210 cells with vitamin K3plus 1 muícysteine 21% of control. These data suggest that GSH depletion may be or 1 HIM reduced glutathione abrogated vitamin remediated a more important component of the antitumor activity of vitamin growth inhibition (Table 2). However, nontoxic concentrations of K3 than are these other quiñones. ascorbic acid and a-tocopherol did not prevent vitamin Ka-medi Coincubation with 1 rriM cysteine restored the vitamin K3- ated growth inhibition. mediated depletion of the L1210 GSH pool (Table 1); coincuba- Superoxide Generation. Ohnishi ef al. have shown that mi- tion with 1 DIM reduced glutathione partially restored the GSH crosomes plus NAD(P)H may reduce vitamin K3 to a semiqui- pool. However, reduced glutathione was somewhat unstable in none, which has a short lifetime and may produce Superoxide aerobic tissue culture medium with up to 43% oxidation of (9,10). In some cells, Superoxide generation has been correlated glutathione occurring in 30 min in medium in the absence of cells, with GSH depletion (27,28,29). In L1210 cells, vitamin Kgcaused with or without the presence of vitamin Kg. Coincubation with almost a 400% stimulation of Superoxide detected extracellularly

CANCER RESEARCH VOL. 45 NOVEMBER 1985 5259

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1985 American Association for Cancer Research. MODULATION OF REDUCED GLUTATHIONE BY VITAMIN

Table 2 Table 3 Effect of reducing agents on vitamin K3-mediatedinhibition of L1210 growth Effect of vitamin K,-reducing agents on NADPH/NADP+poolsin L1210 cells L1210 cells were grown in liquid suspension culture in the continuous presence From 10 to 20 millionL1210 cells per data point were incubated with the agents of 36 MMvitamin K3±reducing agent (see "Materials and Methods'). Shown are shown for 16 h at 37°Cin a 5% CO2:95%air atmosphere, after which pyridine the means ±SD of the percentage of control (i.e., no vitamin Ks exposure) cell nucleotides were measured by the thiazolyl blue cycling assay of Bemofsky and count at 96 h of a representativeexperiment. None of these reducing agents alone Swan (26). Proteins were assayed by the method of Bradford (21). Shown are the affected L1210 growth at all at the concentrations shown. means ±SEof the combined data from 4-7 separate experiments. % of normal growth8 Reducing agent nmol/mg protein None 10±1 AgentNone Ascorbic acid (0.7 mw) 10 ±0.8 a-Tocopherol (0.2 mm) 20 ±2 ±0.04 ±0.2 Cysteine (1 RIM) 110±10 Vitamin K3 MM 0.2 ±0.04 1 ±0.3 Glutathione(1 DIM) 100±3 VitaminK.JVitamin 18MM27 0.2 ±0.07 1 ±0.4 K;, 0.1 ±0.04s 0.6 ±0.3* * Cell count of control cells at 96 h was 4.2 ±0.4 million. MM Vitamin K, 36MM27/iM:l <0.01 <0.1 Vitamin K3:cysteine mu 0.2 ±0.07 1 ±0.2 Vitamin K3:ascorbicacid 27 I/M:1 rriM 0.02 ±0.01 <0.1 Vitamin K3:,,-tocopherolConcentration927 MM:0.2rriMNADP*0.2<0.01NADPH1 <0.1 ' Significantly different than control (P < 0.05 by paired r test).

.£ I4.0- of NADPH and total NADP caused by 27 MMvitamin KS;coincu Q> bation with 0.2 rriMa-tocopherol or 1 mw ascorbic acid did not O o. (Table 3). e I2.0H The ability of vitamin KSto deplete the L1210 cellular NADPH pool was compared to that of several other quiñones.Exposure for 1 h to equitoxic concentrations of the naphthoquinones vitamin «3(72MM),lapacho! (53 MM),dichlorolapachol (53 MM), and phylloquinone (670 MM)diminishedthe NADPH pool to 30 ± 10, 60 ±10, 60 ±11, and 40 ±9% of control (n = 3), 8.0H respectively. The equitoxic concentration of the benzanthroqui-

4 none doxorubicin (37 MM)lowered the NADPH pool to 80 ±12% (C of control. 6. OH DISCUSSION 4. OH u. O The GSH pool is an important determinant of tumor cell viability (30-32). Exposure of leukemia L1210 cells by vitamin Kscauses « 2.0- marked alterations in the GSH pool; the concentration effect 36 0 3.6 18 27 36 45 54 72 MM)cause marked depletion of the GSH pool. Total glutathione is also increased by a low concentration of vitamin KSto a further degree than is GSH and is unaffected by higher vitamin Ks Chart 4. Superoxide generation by 1 to 5 millionL1210 cells in the presence of varying concentrations of vitamin K3(O) or vitamin K3+ 0.2 mw ,,-tocopherol (A) concentrations. These data suggest that vitamin Ks has two was assayed by the reduction of cytochrome c in the presence and absence of effects on thiol metabolism in L1210 cells: (a) to stimulate new SOD, by the method of Smith eíal.(12). Points, means (n = 3); bars, SD. glutathione synthesis; and (b) cause a shift in the GSH:GSSG ratio favoring GSSG. At low concentrations of vitamin KS,new (Chart 4). Superoxide production at 30 min was maximally stim thiol synthesis predominates; thus both GSH and GSSG pools ulated by 45 UMvitamin K:tand did not further increase at higher are raised. At higher concentrations, thiol synthesis cannot keep vitamin Ks concentrations (Chart 4). a-Tocopherol inhibited de up with rapid oxidation to GSSG; thus the GSH pool is depleted. tectable Superoxide generated in the presence of 36 #jMvitamin In L1210 leukemia, vitamin Ks depletes the GSH pool to a Ks by 43% (Chart 4). Meaningful measurements of Superoxide significantly greater degree than do equitoxic concentrations of production in the presence of ascorbic acid or cysteine were its analogues phylloquinone, lapacho!, and dichlorolapachol or precluded by the rapid reduction of ferricytochrome c by these the benzanthroquinone doxorubicin. Our data strongly suggest compounds. that vitamin K3-mediateddepletion of the GSH pool is intimately NADP+:NADPHPools. Incubation of L1210 cells with vitamin associated with its tumor cell growth-inhibitory activity: (a) the Ks for 16 h caused a concentration-dependent depletion of concentrations of vitamin KS required to inhibit L1210 growth NADPH pool with significant depletion of NADPH and total NADP and to deplete the L1210 GSH pool are similar; (b) normal L1210 observed after exposure to vitamin KSin concentrations as low growth in the presence of vitamin «3canbe completely restored as 27 //M(Table 3). From 40 to 50% depletion of NADPHoccurred by coincubation with thiols, such as cysteine or glutathione. It is in the presence of 27 UMvitamin K,, the concentration below unclear yet whether cysteine protects L1210 cells by serving as which growth inhibition and GSH depletion occurred (Charts 2 a substitute nucleophile or by stimulating new glutathione syn and 3). Coincubation with 1 HIMcysteine prevented the depletion thesis. However, other nucleophiles such as ascorbic acid and

CANCER RESEARCH VOL. 45 NOVEMBER 1985 5260

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1985 American Association for Cancer Research. MODULATION OF REDUCED GLUTATHIONE BY VITAMIN K3 a-tocopherol neither prevented vitamin Ka-mediated GSH deple which stimulated Superoxide generation (18-27 /¿M)because tion nor protected L1210 cells from vitamin K3 toxicity. enhanced new glutathione synthesis more than compensated for It is of interest that reduced glutathione protected the cells depletion of GSH by oxidation. The ability of a-tocopherol to from vitamin K3 toxicity in that, unlike cysteine, the tripeptide inhibit vitamin Ka-mediated intracellular Superoxide generation glutathione does not stimulate the L1210 GSH pool (Table 1). without affecting GSH depletion is not consistent with this hy This suggests the possibility that vitamin Ka-mediated cytotox- pothesis and is as yet unexplained. The caution, however, that icity may occur at or near the cell surface, in proximity to the extracellular assay system detects only a small fraction of extracellular glutathione. Morrison eíal. have made a similar Superoxide generated and that a number of transformation steps suggestion in hepatocytes (33). Although we cannot entirely occur before intracellular Superoxide is detected extracellulariy exclude the possibility that glutathione inhibits entry of vitamin must be reemphasized with respect to overinterpretation of the Ka into the cell by covalently forming vitamin K3:glutathione quantitative effect of a-tocopherol on Superoxide generation. thioether adducts, the fact that intracellular GSH pool depletion We observed that vitamin Ka caused partial depletion of the occurs in the presence of the vitamin K3-reduced glutathione L1210 cellular NADPH and total NADP pools at concentrations combination (Table 1) suggests that the cell is subject to some slightly lower than those which disturbed cell growth and the vitamin «3effect, despite the presence of reduced glutathione. GSH pool. Depletion of total NADP suggests that exposure to Further work on the site of vitamin Katoxicity is clearly necessary. vitamin Ka is associated with either inhibition of NADP synthesis There are several mechanisms by which cellular metabolism or enhanced degradation, possibly as a consequence of vitamin of vitamin «3may cause a shift in the GSH:GSSG ratio favoring Ka-mediated damage to macromolecules (33). Reversal of vita GSSG: (a) vitamin K:( semiquinone radical may reduce molecular min Ka-mediated total NADP pool depletion by cysteine, which oxygen, causing Superoxide and hydrogen peroxide formation also prevents vitamin K3 cytotoxicity, but not by other reducing (10, 12, 13, 16). Peroxide generation could consume reduced agents which do not prevent vitamin Ka cytotoxicity is consistent glutathione by the glutathione peroxidase reaction; (ó)reduction with this hypothesis. It must be emphasized that depletion of the of vitamin KS utilizes NAD(P)H cofactor (15, 27), causing a L1210 cellular NADPH pool by vitamin K3>unlike the situation in depletion of the NAD(P)H pool. NADPH is also a cofactor for hepatocytes (13), is not simply due to oxidation of NADPH in the reduction of oxidized glutathione (28, 34); loss of NADPH could presence of vitamin Ka; altered NADP turnover must be involved, impair glutathione reduction; (c) vitamin «3may bind reduced since both NADPH and NADP+ are depleted. glutathione covalently at position 3 via thio ether formation (16). It is difficult to ascertain from these data whether vitamin Ka Since the concentration of GSH in L1210 is much greater than toxicity versus leukemia L1210 was related to GSH depletion, or 100 /tw and we noted GSH depletion at vitamin Kaconcentrations both, since the vitamin KSconcentration dependence of growth of 36 to 54 /¿M,it is stoichiometrically unlikely that covalent inhibition, GSH depletion, and NADPH depletion were similar and binding is a major cause of vitamin K3-rnediated GSH depletion. all of the vitamin «3effects were abrogated by cysteine. If vitamin In L1210 leukemia, vitamin Ka causes rapid Superoxide gen Ka toxicity was related to NADPH depletion, there may be a eration, probably due to reaction of molecular oxygen with the threshold effect, since some depletion of NADPH was observed vitamin K, semiquinone radical. The rate of detectable superox- at a nontoxic vitamin KSconcentration (27 ^M). ide generation is maximal at 45 /UMvitamin Ka and does not The observation that equitoxic concentrations of every qui- increase further at higher vitamin Ka concentrations, suggesting none studied caused significant depletion of the NADPH pool in that 45 fíMvitamin K3 causes the reductase(s) of L1210 respon L1210 cells, albeit to varying degrees, suggests that NADPH sible for the reduction of vitamin Kato its semiquinone to catalyze depletion may be a common cytotoxic effect of the quinone at Vâ„¢,.Therefore a rough estimate for KMof vitamin Ka-mediated drugs. The relative contribution of NADPH depletion to doxorub- Superoxide generation by L1210 cells is 22 ^M, considerably icin-mediated cytotoxicity is probably less than for the naphtho- higher than the reported KM value of vitamin KS of 2 pM for quinones, however, since a toxic concentration of doxorubicin NADPH oxidation by liver microsomes (11). However, since only caused only minimal NADPH pool depletion. 3% of Superoxide generated intracellularly is detected by reduc In contrast, only vitamin «3caused significant GSH pool deple tion of cytochrome c extracellulariy (29), only limited inferences tion. It has been suggested that GSH depletion may be an pertaining to the rate of Superoxide generation can be made. important mechanism of toxicity of other antitumor quiñones, The vitamin K3 concentration dependence of stimulation of such as doxorubicin (35-37). Doroshow ef al. (38) noted that Superoxide generation did not correlate well with the vitamin Ka doxorubicin administration in vivo to mice moderately depleted concentration dependence of GSH depletion, especially in the hepatic and cardiac GSH pools. However, our data and those of crucial concentration range 18-45 ftw. The good correlation of others suggest that GSH depletion is not a general mechanism growth inhibition with GSH depletion and lack of correlation of of toxicity of all antitumor quiñones.Patterson ef al. (39) observed growth inhibition with Superoxide generation suggest that growth GSH depletion in rat heart due to some antitumor benzanthro- inhibition was related to the former rather than to the latter. quinones (e.g., daunorubicin), but not others (e.g., 1,4-bis¡2-[(2- However, the vitamin KS concentration dependence of GSSG hydroxyethyl)amino]ethylamino|-9,10-anthracenedione). We also generation, as estimated from the difference between the total noted a disparity among quiñones in their effect on the GSH glutathione pool (expressed as percentage of control) and the pool; vitamin Ka caused GSH pool depletion, whereas lapacho!, GSH pool (also expressed as percentage of control; see Chart dichlorolapachol, and phylloquinone did not. Also unlike its effect 2), correlated well with the vitamin Ka concentration dependence on mouse heart and liver, a toxic concentration of doxorubicin of the Superoxide generation rate. This suggests that oxidation did not affect the GSH pool in cultured L1210 cells. of GSH was related to Superoxide production. GSH depletion Since the GSH pool may be an important determinant of was not observed at the lower concentrations of vitamin Ka resistance to some chemotherapeutic agents, e.g., certain alkyl-

CANCER RESEARCH VOL. 45 NOVEMBER 1985 5261

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1985 American Association for Cancer Research. MODULATION OF REDUCED GLUTATHIONE BY VITAMIN «3 ating agents (40), as well as a mediator of antitumor activity, the Biochem. Biophys., 224: 568-578,1983. 17. Wills, E. D. Effects of vitamin K and naphthoqumones on hpid peroxidase identification of agents which selectively deplete GSH has be formation and oxidative demethylation by liver microsomes. Biochem. Phar- come an important research goal. Vitamin Ka, which is among macd.. 27:1879-1883,1972. the most potent GSH-depleting agents of the antitumor qui 18. Block, J. B., Isacoff, M., and Dietrich, M. Amphotericin B effects in tissue culture cell lines. Cancer Treat. Rep., 63: 643-645,1979. ñones,may be a useful template for modeling of GSH-depleting 19. Dulbecco, R., and Vogt, M. Plaque formation and isolation of pure lines with agents with improved therapeutic:toxic ratios. poliomyelitis viruses. J. Exp. Med., 99: 167-182,1954. 20. Beutler, E., Duron, O., and Kelly. B. M. Improved method for the determination of blood glutathione. J. Lab. Clin. Med., 67: 882-888.1982. 21. Bradford, M. A rapid and sensitive method for the quantitatton of microgram ACKNOWLEDGMENTS quantities of protein utilizing the principle of protein dye binding. Anal. Biochem., 72: 248-254,1976. We would like to thank Dr. James L. Doroshow, City of Hope Medical Center, 22. Cadman. E. C., Heimer, R., and Benz, C. The influence of methotrexate for helpful discussions and Christie Seltzer and Barbara Young for secretarial pretreatment on 5-fluorouracil metabolism in L1210 cells. J. Bid. Chem., 256: assistance. 1695-1704,1981. 23. Wohlhueter, R. M., Marz, R., Graft, J. D., and Plagemann, P. G. W. A rapid mixing technique to measure transport in suspended animal cells: application REFERENCES to nucleoside transport in Novikoff rat hepatoma cells. Methods Cell Bid., 20: 211-236,1978. 1. Chtebowski, R. T., Block, J. B., and Dietrich, M. F. Vitamin K effects on 24. Akerboom, T. P. M., and Síes,H. Assay of glutathione, glutathione disulfide, cytotoxicity and biosyntheiic pathway inhibition in cancer. In: f. L. Meyskens and glutathione mixed disulfides in biological samples. Methods Enzymol,, 77: and K. Prasad (eds.), Modulation and Mediation of Cancer by Vitamins, pp. 373-382,1981. 226-281. Basel: S. Karger AG. 1983. 25. Massey, V. The microestimation of Buccinate and the extinction coefficient of 2. Chtebowski. R. T., Block, J. B., Dietrich, M., Octay, E.. Barth. N., Yanagihara, cytochrome c. Biochim. Biophys. Acta, 34:255-256,1959. R., Gota, C., and Ali, I. Inhibition of human tumor growth and DMA biosynthetic 26. Bemofsky, C., and Swan, M. An improved cycling assay for nicotinamide activity by vitamin K and warfarin; m vitro and clinical results. Proc. Am. Assoc. adenine dinucleolide. Anal. Biochem., 53: 452-458,1973. Cancer Res., 24: 653,1983. 27. Rubin, R. P. Calcium in the Secretory Process. New York: Plenum Publishing 3. Chtebowski, R. T., Dietrich, M., Akman, S. A., and Block, J. B. Vitamin K Corp., 1974. (menadione) inhibition of human tumor growth in the soft agar assay system. 28. Kosower, N. D., and Kosower, E. M. The glutathione status of cads. Int. Rev. Proc. Am. Soc. Clin. Oncol.. 3:93,1984. Cytd., 54:103-106,1978. 4. Chtebowski, R. T., Dietrich. M., and Block, J. Warfarin plus vitamin K enhance 29. POWIS, G., Svingen, B. A., and Appel, P. Quinone-stimulated Superoxide methotrexate and 5-fluorouracil effects in L1210 cells. Clin. Res., 29: 63A, formation by subcellular fractions, isolated hepatocytes and other cells. Md. 1981. Pharmacd., 20: 387-394,1981. 5. Harrison, P. R. Inhibition by ATP of the growth-inhibitory effect of Synkavite 30. Arrick, B. A., Nathan, C. F., and Cohn, Z. A. Inhibition of glutathione synthesis (2-methyi-1.4-naphthoqumone bis disodium phosphate) on mouse ascites tu augments lysis of murine tumor cells by sulfhydryl-reactive antineodastjcs. J. mor cells. Br. J. Cancer, 24:807-816,1970. Clin. Invest., 77: 255-267,1983. 6. Prasad, K. N., Edwards-Prasad. J., and Sakamoto, A. Vitamin K3 (menadione) 31. Arrick, B. A., Nathan, C. F., Griffith, O. W., and Cohn, Z. A. Glutathione inhibits the growth of mammalian tumor ceils in culture. Life Sci., 29: 1387- depletion sensitizes tumor cells to oxidative cytdysis. J. Bid. Chem., 257: 1392,1981. 766-782,1981. 7. Shoemaker. R. H.. Wolpert. J. K., Makuch, R. W., and Venditi!. S. M. Use of 32. Nathan, C. F., Arrick, B. A., Murray, H. W., DeSantjs, N. M., and Cohn, Z. A. the human tumor clonogenic assay for new drug screening. Proc. Am. Assoc. Tumor cell antioxidant defenses. Inhibition of the glutathione redox cycle Cancer Res., 24:311,1983. enhances macrophage-mediated cytdysis. J. Exp. Med., 753:766-782,1981. 8. Lmd, C., Höchstem,P., and Ernster, L. DT diaphorase as a quinone reducíase: 33. Morrison, H., Hernstrom, B., Nordenskjdd, M., Thor, H., and Orrenius, S. a cellular control device against semiquinone and Superoxide radical formation. Induction of DNA damage by menadione (2-methyl-1-4-naphthoqumone) in Arch. Biochem Biophys., 216: 175-185,1982. primary cultures of rat hepatocytes. Biochem. Pharmacd., 33: 1763-1769, 9. Nakamura. S., and Yamazaki, I. A mixed mechanism in the reaction of mild 1984. xanthme oxidase with electron acceptors. Biochim, Biophys. Acta, 789: 29- 34. Flohe, L., Benohr, H., Síes,H., Wally, D., and Wendel, A. Glutathione. New 37, 1969. York: Academic Press, 1974. 10. Ohnishi. T., Yamazaki, H., Inanagi, T., Nakamura, T.. and Yamazaki, I. One 35. Babson, J. R., Abe«,N. S., and Reed, D. J. Protective rote of the glutathione electron transfer reactions in biochemical systems. On the reaction of free redox cycle against Adriamycin-mediated toxicity in isdated hepatocytes. radicals formed in the enzymatic oxidation. Biochim. Biophys. Acta, 189: 29- Biochem. Pharmacd., 30: 2299-2304,1981. 37,1969. 36. Meredith, E. J., and Reed, D. J. Depletion in vitro of mitochondria! glutathione 11. Sato, R., Nishibayashi, H., and Omura, T. Naphthoquinone-dependent oxida in rat hepatocytes and enhancement of lipid peroxidation by adriamycin and tion of reduced triphosphopyridine nucteotide by liver microsomes. Biochim. 1,3 bis(2-chloroethyl)-1-nitrosourea (BCNU). Biochem Pharmacd., 32: 1338- Biophys. Acta, 63: 550-552,1962. 1383,1983. 12. Smith, M. T., Thor, H., and Oremus, S. Detection and measurement of drug- 37. Minnaugh, E. G., Cram, J. E., and Truch, M. A. Stimulation of mouse heart induced oxygen radical formation. Methods Enzymol.. 705: 505-510,1984. and liver microsomal lipid peroxidation by anthracydine anticancer drugs; 13. Thor, H., Smith, M., Hartzeii. P., Bellomo, G., Jewell, S., and Orrenius, S. The characterization and effects of reactive oxygen scavengers. J. Pharmacd. metabolism of menadione (2-methyl-1,4-naphthoquinone) by isolated hepato- Exp. Ther., 226: 806-816,1983. cytes. J. Bid. Chem., 257:12419-12425,1982. 38. Doroshow, J. H., Locker, G. Y., BakJinger, J., and Myers, C. E. The effect of 14. Gallin, J. I., Seligmann, B. E., Cramer, E. B., Schiffmann, E., and Fletcher, M. doxorubicm on hepatic and cardiac glutathione. Res. Commun. Chem. Pathd. P. Effects of vitamin K on human neutrophil function. J. Immunoi.. 728:1399- Pharmacd., 26: 285-295,1979. 1408,1982. 39. Patterson, L. H., Gandecha, B. M., and Brown, J. R. 1,4 Bis(2-((2-hydroxy- 15. Bellomo, G., Jewell, S. A., Thor, H., and Orrenius, S. Regulation of intracellular ethyl)ammo)ethyl amino)-9,10 anthracenedione, an anthraquinone antitumor calcium compartmentation studies with isolated hepatocytes and ( butyl-hydro- agent that does not cause lipid peroxidation in vivo; comparison with dauno- peroxide. Proc. Nati. Acad. Set. USA, 79: 6842-6846,1982. rubicm, Biochem. Biophys. Res. Commun., 770: 399-405,1983. 16. Wefers, H., and Síes.H. Hepatic low-level chemiluminescence during redox 40. Somfai-Relle. S., Suzukake, K., Vistica. B. P., and Vistica. D. T. Glutathione- cycling of menadione and the menadione-glutathione conjugate: relation to conferred resistance to antineoplastics: approaches toward its reduction. glutathione and NAD(P)H: quinone reducíase (DT diaphorase) activity. Arch. Cancer Treat. Rev., 77:43-54,1984.

CANCER RESEARCH VOL. 45 NOVEMBER 1985 5262

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1985 American Association for Cancer Research. Modulation of Cytotoxicity of Menadione Sodium Bisulfite versus Leukemia L1210 by the Acid-soluble Thiol Pool

Steven A. Akman, Marian Dietrich, Rowan Chlebowski, et al.

Cancer Res 1985;45:5257-5262.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/45/11_Part_1/5257

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/45/11_Part_1/5257. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1985 American Association for Cancer Research.