Cooxidation of Cyclophosphamide As an Alternative Pathway for Its Bioactivation and Lung Toxicity1

[CANCER RESEARCH 51. 542-548. Januar> 15. 1991]

Cooxidation of Cyclophosphamide as an Alternative Pathway for Its Bioactivation and Lung Toxicity1

Robin D. Smith2 and James P. Kehrer' Division of Pharmacology and Toxicology, College of Pharmacy, The I'nirersily of Texas at Austin, Austin, Texas 78712-1074

ABSTRACT actions, has been reported after single doses in experimental animals and humans (2, 3). In rats, the damage appears as a A single i.p. dose of cyclophosphamidc produces lung cell injury and fibrosi* in mice. Although cyclophosphamidc is activated by the cyto- focal sclerosing pneumonitis at the alveolar level, with injury chrome P-450 mixed function oxidase (MKO) system, a role for this to type I cells, abnormal type II cells, a gradually increasing system in the development of lung injury has not been established. The content of collagen characteristic of fibrosis, and areas of alveo involvement of other metabolic pathways, such as cooxidation via pros- lar collapse (4). The lungs of mice given a single injection of taglandin II synthase, in the toxicity of Cyclophosphamide has not been Cyclophosphamide showed similar histopathology (5, 6). The studied. The objectives of the current study were to assess the effects of time course of the acute lung lesion in mice is relatively rapid, various inhibitors of MIO and prostaglandin II synthase activity on the with the alveolar labeling index being maximal after 5 days and development of cyclophosphamide-induced lung damage and fibrosis in pulmonary thymidine incorporation reaching a peak about 1 mice, to determine whether arachidonic acid as well as NADPII could week after treatment with a single dose (7, 8). support the activation of Cyclophosphamide to an alkylating metabolite, The mechanism of cyclophosphamide-induced lung damage and to assess the capacity of Cyclophosphamide to serve as a reducing has not been established, although the well described hepatic cosubstrate. In addition, the ability of a low dose of Cyclophosphamide metabolism of this drug by MFO4 enzymes to alkylating and to prevent the lung injury from a later higher dose was determined. Treatment with SKK 525A, piperonyl butoxide, or 1-benzylimidazole, oxidizing metabolites (9-11) has led to the assumption that this followed by a single 200 mg/kg dose of Cyclophosphamide, did not enzyme system is required for both the therapeutic and toxic diminish pulmonary thymidine incorporation (an index of cell division effects of Cyclophosphamide. Hyperoxia aggravates the cyclo after injury) or hydroxyproline content (an indicator of fibrosis), com phosphamide-induced lung lesion (7) and Patel (12) has sug pared to mice treated with Cyclophosphamide alone. Pretreatment with gested that oxidative injury is involved in the damaging process. l-aminobenzotriazole reduced the incorporation of thymidine into lung However, although metabolic activation must occur before cy- DNA on days 3 and 10, but not on day 7, and also reduced lung totoxic effects develop and MFOs can metabolize Cyclophos hydroxyproline accumulation. Treatment with indomethacin, nordihydro- phamide to alkylating metabolites in vitro, it is unclear what guiaretic acid, or aspirin prior to Cyclophosphamide greatly reduced levels role this particular enzyme system plays in the in vivo effects of of pulmonary thymidine incorporation and/or hydroxyproline content, compared to Cyclophosphamide alone. Low dose pretreatment with Cyclo this drug. MFO-mediated metabolism is an important, but not exclu phosphamide did not prevent the lung injury or fibrosis from a subsequent higher dose. NADPII supported greater production of alkylating metab sive, pathway to bioactivate various xenobiotics. This is illus olites in liver than in lung microsomes. In contrast, the arachidonic acid- trated by the increasing recognition of the role hydroperoxide- dependent oxidation (termed "cooxidation") plays in xenobiotic supported production of alkylating metabolites was greater in lung mi crosomes. No NADPII- or arachidonate-supported alkylating activity metabolism and subsequent carcinogenesis or tissue injury (13). was evident in lung or liver cytosol. SKI 525A and 1-aminobcnzotriazole The oxidation of various chemicals during the metabolism of inhibited the NADPII-supported reaction in liver, but not lung, while arachidonic acid by PHS was first shown in 1975 (14). This indomethacin and nordihydroguiaretic acid inhibited the arachidonic acid- was subsequently determined to involve the peroxidase portion supported reaction in lung but not liver. Cyclophosphamide was a mod erately active reducing cosubstrate for 5-phenyl-4-pentenyl hydroperox- of PHS, which reduces the prostaglandin G? formed by the cyclooxygenase reaction to the alcohol prostaglandin H2 (15). ide in both lung and liver microsomes. These results demonstrate that Numerous compounds which are structurally unrelated to fatty- pathways in lung tissue unrelated to MKOs can metabolize Cyclophos phamide to an alkylating compound and that MFO-mediated activation acid endoperoxides can be oxidized during this reduction step of Cyclophosphamide may not be essential for the development of the (13). Additional studies have shown that similar reactions can pulmonary toxicity associated with this drug. be catalyzed by lipoxygenase activities. Evidence is available supporting the concept that cooxidation INTRODUCTION by PHS, together with metabolism by pulmonary MFO sys tems, mediates the lung toxicity of 3-methylindole (16). Lung A variety of chemical compounds, including numerous cyto- tissue is capable of metabolizing Cyclophosphamide to alkylat toxic drugs, are known to damage lung tissue following systemic ing species (17), and similar metabolic pathways could be administration (1). Lung damage due to antineoplastic drugs involved in the pulmonary toxicity of this drug. Interestingly, can be produced by a number of factors, including direct drug in contrast to MFOs, which are found in the highest concentra action or the synergistic or antagonistic action of two drugs or tions in the liver, PHS and lipoxygenase activities are relatively of the same drug with itself. The lung damage produced by high in lung and bladder (13), sites of the major cyclophospha Cyclophosphamide, although possibly modified by such inter- mide-induced toxicities. The present study provides evidence Received8/6/90;accepted10/25/90. that the lung damage and fibrosis due to Cyclophosphamide is The costs of publication of this article were defrayed in part by the payment not decreased by most inhibitors of MFO activity or by a of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 4The abbreviations used are: MFO. mixed function oxidase; PHS. prostaglan 1This work was supported by Grant HL 35689 from the National Heart. Lung din H synthase; ABT. l-aminoben/otria/ole: PB\. piperonyl butoxide: Bl. 1- and Blood Institute. ben/ylimidazole: NDGA. nordihydroguiaretic acid; BUT. butylaled hydroxyto- 2Current address: CH2M Hill. 625 Herndon Parkway. Herndon, VA. luene; BUA. butylated hydroxyanisole; NBP. 4-(/>-nitroben*yl)pyridine: PPHP. ' To whom requests for reprints should be addressed. 5-phenyl-4-pentcnyl hydroperoxide: PPA. 5-phenyl-4-pentenyl alcohol.

542

preliminary dose of this drug, which previous work suggested resultant supernatants at 100,000 x g for 60 min. The final supernal ani s decreased the lung toxicity of a subsequent higher dose (5). In were saved on ice and the microsomal pellets were washed in 0.1 M contrast, lung cell damage and pulmonary fibrosis are signifi potassium pyrophosphate, pH 7.4, containing 1 misi EDTA and 20 MM cantly reduced by inhibitors of PHS. The in vitro production of BHT. The final microsomal pellet was resuspended in 1.15% KCl. All alkylating metabolites was primarily NADPH dependent in microsomal and cytosolic fractions were prepared immediately before liver microsomes and primarily arachidonic acid dependent in their use to avoid potential problems with enzyme stability. Protein determinations were by the microbiuret method (25). lung microsomes. It was also found that cyclophosphamide could function in the presence of lung or liver microsomes, as Measurements of Alkylating Metabolites. The conversion of cyclo phosphamide to metabolites capable of alkylating NBP was determined a moderately active reducing cosubstrate. in both microsomes and cytosol. The alkylation of NBP by nitrogen mustards results in a compound which forms a chromophore under MATERIALS AND METHODS alkaline conditions, with an absorbance maximum of 540 nm (26). The standard reaction mixture contained 9 /jmol cyclophosphamide, 2 ^mol Animals and Treatments. Male ICR mice (24-32 g, 6-IO weeks of NADPH, 20 /Â¿molglucose-6-phosphate, 200 MHIO!K:HPO4 (pH 7.4), age) were used for all studies. The animals were obtained from HarÃan 10 /Â¿molMgCb, 2 mg microsomal or cytosolic protein, and sufficient Sprague Dawley (Houston. TX) and had free access to standard lab 1.15% KCl to make the final volume 3 ml. The mixture was preincu- chow and water. Cyclophosphamide (50 or 100 mg/kg), SKF 525A (10 bated at 37Â°Cfor 5 min and the reaction was started by adding 2 units or 50 mg/kg), ABT (50 mg/kg), or aspirin (50 mg/kg) were adminis glucose-6-phosphate dehydrogenase. In other studies, the NADPH- tered i.p. as saline solutions. PBX (400 mg/kg) was given as a single generating system was replaced by 300 nmol arachidonic acid, which i.p. dose in corn oil. BI (25 mg/kg) was administered as a single i.p. was added to initiate the reaction. Inhibitors were added to the reaction dose in 95% ethanol. Inhibitors of MFO activity and aspirin were given mixture (100 MM)and allowed to incubate for 5 min prior to the as single i.p. injections 2 h prior to cyclophosphamide. Indomethacin addition of cyclophosphamide. (1.5 mg/kg in corn oil) was administered twice daily for 2 days and The metabolism of cyclophosphamide was terminated after 15 min again 2 h prior to cyclophosphamide. NDGA was dissolved in 0.1 N by adding 1 ml of a 5.5% (w/v) aqueous ZnSOj-7H2O solution, NaOH and diluted with an equal volume of saline, and the pH was followed by 1 ml of 4% (w/v) aqueous Ba(OH)2-8H2O. This mixture adjusted to ~8.0 with 0.1 N HC1. The resulting solution was adminis was centrifuged at ~1000 x g for 5 min, and the resulting supernatant tered i.p. (25 mg/kg) 2 h prior to cyclophosphamide and at the same was mixed with 1 ml of 0.2 M sodium acetate buffer, pH 4.0. The pH time on days 1 and 2 following cyclophosphamide. The concentrations of this mixture was adjusted to exactly 4.0 to promote maximal color of all solutions were adjusted so that mice received 0.1 ml/10 g body development. NBP (0.4 ml of a 50 mg/ml solution in acetone) was then weight. Control animals received equal volumes of saline. All doses and added and the tubes were heated at 100'C for 20 min. Upon removal treatment schedules were selected based on previous studies showing from the water bath, each tube was shaken vigorously and allowed to an effect on lung enzyme activity or lung toxicity of other xenobiotics (18-21). Mice were euthanized by cervical dislocation. cool. Seven ml of a 5:2 (v/v) ethyl acetate:acetone mixture was then Chemicals. Cyclophosphamide (Cytoxan) was obtained from Bristol- added, followed by 1.5 ml of 5 N NaOH. The mixture was shaken and Myers (New York). Radiolabeled [2-'4C]thymidine (53 Ci/mol) was the absorbance of 3 ml of the upper organic phase was measured at 540 obtained from Amersham (Arlington Heights, IL). BI, 4-nitrobenzyl nm. A standard curve was generated by using known amounts of alcohol, and L-hydroxyproline were obtained from Aldrich Chemical mechlorethamine hydrochloride. All absorbance readings were made Co. (Milwaukee, WI). Triethanolamine hydrochloride, glucose-6-phos- exactly 5 min after the addition of NaOH under minimal light, because phate (disodium salt), BHT, BHA, glucose-6-phosphate dehydrogenase, the chromophore is unstable. arachidonic acid, mechlorethamine, NBP, aspirin, indomethacin, and Reducing Cosubstrate Assay. The ability of cyclophosphamide to NDGA were obtained from Sigma Chemical Co. (St. Louis, MO). sene as a reducing cosubstrate for cytosolic and microsomal peroxidase PPHP was obtained from Oxford BiomÃ©dicalResearch (Oxford, MI). activities was determined by measuring the reduction of PPHP to PPA Piperonyl butoxide was obtained from Pfaltz and Bauer, Inc. (Stamford, (27). Microsomes from lung or liver were preincubated at 37Â°Cfor 3 CT). ABT was a generous gift from Dr. Alan Buckpitt of the University min in 2 ml of 100 mivisodium phosphate, pH 7.4, containing 200 MM of California, Davis. SKF 525A was a gift from Smith-Kline Beecham. Tween 20 and 100 MMcyclophosphamide. PPHP (50 MM)was added All other chemicals and solvents were of the highest purity available. to initiate the reaction. The amount of protein used was that required Pulmonary DNA Synthesis and Hydroxyproline Analyses. DNA syn to reduce 50% of PPHP to PPA in the presence of 100 MMBHA. thesis was assessed by measuring the incorporation of radiolabeled Cytosolic fractions were tested using 0.5 mg protein. Indomethacin, thymidine into total lung DNA. Mice were given an i.p. injection of NDGA, or ABT were added at a concentration of 200 MMand allowed 0.5 nCi [MC]thymidine. After 90 min, animals were killed and the to preincubate for 3 min prior to the addition of cyclophosphamide (6 specific activity of pulmonary' DNA was determined (22). Total lung min prior to PPHP). collagen was estimated by measuring hydroxyproline, an amino acid Reactions were terminated 3 min after the addition of PPHP by solid found primarily in collagen (23). Lung tissue was excised, rinsed in phase extraction on SEP-PAK (Rainin Instruments, Woburn, MA) CIS water, and lyophilized intact. The entire lung was hydrolyzed in 6 N reverse phase cartridges. PPHP and PPA were eluted with 2 ml meth- HC1 and the hydrolysate was assayed for hydroxyproline by oxidation anol and the internal standard, p-nitrobenzyl alcohol (50 MM),was to pyrrole with chloramine T and measurement at 560 nm of the absorbance of the chromophore formed by the reaction with p-dimeth- added. The eluate was injected directly onto a C8 reverse phase column and eluted Â¡socraticallywith a mobile phase of 65% (v/v) metha- ylaminobenzaldehyde (24). Preparation of Microsomes and Cytosol. Pulmonary and hepatic mi nolrwater, at a flow rate of 1.0 ml/min. Detection was at 254 nm. Under these conditions, p-nitrobenzyl alcohol, PPA, and PPHP had crosomes were prepared by standard differential centrifugation proce dures. Tissues were perfused free of blood with ice-cold 1.15% KC1. retention times of 4.2, 9.4, and 10.7 min, respectively. The amount of Lungs from three mice were pooled to give sufficient microsomal PPA produced, and the unreacted PPHP, were quantitated by the protein for the assays and homogenized for 20 s with a Tekmar internal standard method (27). Data are expressed as an index number, Tissumizer in 0.1 M Tris-acÃ©tate,pH 7.4, containing 0.1 M KCl, 1 mivi calculated as described previously (15, 27). EDTA. and 20 MMBHT (to prevent lipid peroxidation). The resulting Statistics. All data are expressed as means Â±SE.Multiple group data homogenate was further homogenized with a Teflon-glass homoge- were analyzed by one-way analysis of variance and the method of nizer. Liver tissue from a single mouse was minced with scissors and planned comparisons. Comparisons between groups were done with the homogenized with a Teflon-glass homogenizer. The final lung and liver Student-Newman-Keuls post hoc test (28). A P value of less than 0.05 homogenates were centrifuged at 10,000 x g for 20 min and the was considered significant. 543

RESULTS on day 7 was significantly greater than in those treated with cyclophosphamide alone. All animals treated only with inhibi Pulmonary Thymidine Incorporation after Cyclophospha- iniili1.There were no deaths in any groups treated with cyclo- tors of MFO activity had thymidine incorporation levels which did not differ from those of saline-treated controls (Fig. 1) at phosphamide. In agreement with earlier investigations (7, 8), any time point (data not shown). the levels of thymidine incorporation were maximal 7 days after In contrast to the minimal effects achieved with inhibitors of either 100 or 200 mg/kg cyclophosphamide (Fig. 1). Inhibitors MFO activity, prctrcatment of mice with indomethacin or of MFOs did not alter the peak incorporation of thymidine on NDGA significantly reduced the level of thymidine incorpora day 7 produced by a single dose of 200 mg/kg cyclophospha tion produced by cyclophosphamide on most days studied (Fig. mide (Fig. 2). Thymidine incorporation was also unaffected on 3). While the levels were still elevated above saline-treated days 3 and 10 in all groups, except for a decrease with PBX on controls on day 7, they were reduced 47'r (indomethacin) and day 3 and decreases with ABT on days 3 and 10 (Fig. 2). In 60rf (NDGA) compared to those animals receiving cyclophos those mice receiving pretreatment with BI followed by a single phamide alone (Fig. 1). Mice treated with NDGA plus cyclo dose of cyclophosphamide, the level of thymidine incorporation phosphamide had thymidine incorporation levels which were significantly greater than saline controls only on day 7. Mice given a dose of 50 mg/kg cyclophosphamide followed 6000 . 7 days later by 250 nig/kg had levels and patterns of thymidine incorporation similar to those of animals given only a single dose of 250 mg/kg cyclophosphamide (Fig. 4). A single injec tion of 50 mg/kg cyclophosphamide did not produce significant c- 4000. C O* increases in thymidine incorporation at any time point meas ured. In mice receiving a dose of 100 mg/kg cyclophosphamide 11 on day 0 followed by another 100 mg/kg dose on day 7, a 2000 . second, albeit smaller, peak was seen on day 14, indicating some modest modification of the lung injury or repair process (data not shown). Lung Hydroxyprolinc Content following Treatment with Cy clophosphamide. With the exception of ABT. inhibitors of 0 9 12 15 21 MFO activity were unable to prevent the accumulation of lung Time (days) hydroxyproline that occurred after treatment with cyclophos Fig. I. Effect of a single dose of cyclophosphamide on mouse pulmonary thymidine incorporation. Animals were treated i.p. with 100 (O) or 200 (â€¢)mg/ phamide (Fig. 5). In fact, both PBX- and Bl-pretreated animals kg cyclophosphamide on day 0. C'onlrol mice were given a single injection of had lung hydroxyproline levels which were significantly greater saline (A). Data are expressed as means Â±SE. SE were smaller than the symbol size where error bars are not visible, n = 5-8 for all time points. *. significantly than those in mice treated with cyclophosphamide alone. All different from saline controls at the same time point (/"< 0.05). inhibitors of arachidonic acid metabolism tested significantly inhibited the cyclophosphamide-induced accumulation of hy droxyproline (Fig. 5). The levels of lung hydroxyproline seen were the same as after inhibitor alone except for NDGA, which 8000 . allowed a small but significant increase. Treatment of animals

O 4000 ~t 6000. O

3000 â€¢ 4000. feaIt

li?'Â¿7. 200(1 â€¢ 'â‚¬= | 2000. I 1000-

0 0 3 6 9 12 15 IS 21 Time (days) o Fig. 2. Effects of MFC) inhibitor prctrcatment on pulmonary thymidine incor o 6 9 12 15 IK 21 poration due to a single dose of 200 mg/kg cyclophosphamide. Mice were given Time (days) a single i.p. dose of SKF 525.V (D) (50 mg/kg). piperonyl butoxide (â€¢)(400 mg/ Fig. 3. Effects of arachidonic acid metabolism inhibitors on pulmonary thy- kg), l-bcnzylimidazolc (O) (25 mg/kg). or l-aminobenzotriazole (â€¢)(50 mg/kg) inniiin- incorporation due to a single dose of 200 mg/kg cyclophosphamide. Mice 2 h prior to cyclophosphamide. Thymidine incorporation after treatment with were given pretreatments with indomethacin (D) (1.5 mg/kg twice daily for 2 inhibitors alone was not significantly different from saline controls and is not days and again 2 h before cyclophosphamide) or NDCJA (â€¢)(asingle 25 mg/kg presented. Refer to Fig. 1 for saline control and 200 mg/kg cyclophosphamide dose 2 h before cyclophosphamide). Thymidine incorporation after treatment values. Data are expressed as means Â±SE. SE were smaller than the symbol size with inhibitors alone was not significantly different from saline controls and is where error bars are not visible, n = 5-8 for all time points. *. significantly not presented. Refer to Fig. 1 for saline control and 200 mg/kg cyclophosphamide different from saline control at the same time point (P < 0.05): t. significantly values. Data are expressed as means Â±SE. SE were smaller than the symbol siÂ« greater than 200 mg/kg cyclophosphamide alone at the same time point (P < where error bars are not visible, n = 5-8 for all time points. *. significantly greater 0.05); Â§.significantly less than 200 mg/kg cyclophosphamide alone at the same than saline control at the same time point (P < 0.05): t. significantly less than time point (P < 0.05). 200 mg/kg cyclophosphamide at the same time point (P < 0.05). 544

7000 other determinations. A small amount of NBP-alkylating sub stances were produced in both lung and liver microsomes in cubated with cyclophosphamide in the absence of exogenous NADPH or arachidonic acid (Table 1). Alkylating activity was Â¿Ã®5000 - significantly enhanced in both lung and liver microsomes by 2 the addition of the NADPH-generating system, although the

I 4000 - activity of liver microsomes was approximately twice that of "rs the lung. Both SKF 525A and ABT decreased alkylating activity i S 3000 - in liver microsomes to the levels seen in lung. Neither drug had any effect on the NADPH-stimulated alkylating activity in lung microsomes. An arachidonic acid-stimulated increase in alkylating activity was evident in both lung and liver microsomes (Table 1). 1000 However, in contrast to the NADPH-mediated activity, the lung microsomes exhibited significantly greater alkylating ca pacity than liver. Indomethacin, NDGA, and ABT decreased \s 18 21 lung microsomal alkylating activity to the levels seen in liver Time (days) microsomes but had no effect on the arachidonic acid-stimu Fig. 4. Effect of Â¡ipreliminary dose of cyclophosphamide on pulmonary thymidine incorporation due to a second dose. Mice were given a single dose of lated activity in liver microsomes. 50 mg/kg cyclophosphamide (â€¢)or saline (â€¢)on day -7. followed by 250 mg/kg None of the inhibitors tested decreased NADPH- or arachi cyclophosphamide on day 0. Control mice were gi\en a single dose of 50 mg/kg donic acid-dependent alkylating activity to baseline values, per cyclophosphamide (O| or saline (G) on day â€”7.followed by saline on day 0. Data are expressed as means Â±SE. SE were smaller than the symbol si/e where error haps due to insufficient dosages or to activation by other bars arc not visible. Â«= 5-8 for all time points. *. significantly different from metabolic pathways. Cytosol from either lung or liver did not saline/saline (/>< 0.05).

M500-3Â« aspirin 400-~ |â€”* NDGA 300- indomethacin uo o.g 1-aminoben/ot ria/ole 2Ã•X)-E

UK) - pipcronyl butoxide

l-benzylimidazole 0-*â€” Saline/Saline Saline/CP250 CP50/Saline CP50/CP250 SKF 50 mg/kg Treatment Fig. 6. Effects of a preliminary 50 mg/kg dose of cyclophosphamide (CP) on SKF 10 mg/kg pulmonary hydroxyproline accumulation due to 250 mg/kg cyclophosphamide. Mice were given either saline or a single dose of 50 mg/kg cyclophosphamide on CP 200 mg/kg day -7, followed by a single dose of 250 mg/kg cyclophosphamide on day 0. Controls received an injection of saline or 50 mg/kg cyclophosphamide on day Saline controls -7. followed by saline on day 0. Hydroxyprolinc content was measured on day 28. Data are expressed as means Â±SE. n = 7-10 for all time points. *. significantly 200 400 600 800 different from saline/saline (/>< 0.05). Hydroxyproline (tig/lung) Fig. 5. Effects of inhibitors of MFOs or arachidonic acid metabolism on Table I \ADPH and arachidonic acid-mediated production of \HP-alkylating pulmonary hydroxyproline accumulation in mice after a single dose of 200 mg/ metabolites of cyclophosphamide kg cyclophosphamide (CP). Treatments with inhibitors and cyclophosphamide Two mg microsomal protein were incubated at 37Â°Cin the presence of 3 m\l are the same as described in the legends to Figs. 2 and 3. An additional group of cyclophosphamide and either 0.67 m M NADPH (including a generating system) animals was treated with aspirin (50 mg/kg) 2 h prior to cyclophosphamide. or 100 ^M arachidonic acid. Inhibitors were present at 100 >iM. n = 3 for all Controls received saline or the inhibitor, using identical treatment protocols. groups, except cyclophosphamide alone, where n = 6. Compared to saline controls, treatments with inhibitors alone had no significant effects on hydroxyproline content and are not shown. Hydroxyproline content Production of metabolites was measured on day 28. Data are expressed as means Â±SE. n = 5-8 for all time (nmol/15 min/mg)" points. *. significantly different from 200 mg/kg cyclophosphamide alone (P < 0.05): t. significantly different from saline controls (P < 0.05). mi mi- TreatmentCvclophosphamide crosomes19Â± crosomcs22

alone 1 Â± 1 with inhibitors alone had no significant effects on lung hydrox +NADPH 240 + 6 128 Â±7* 237 Â±7*106 yproline content, compared to saline (data not shown). acid+NADPH+Arachidonic 129Â±7110Â±8f A 50 mg/kg dose of cyclophosphamide did not protect against + SKF 525A Â±4 102 Â±7''98 the development of lung fibrosis due to a 250 mg/kg dose given +NADPHABT+ + 133 Â±12116Â± 7 days later (Fig. 6). A single dose of 50 mg/kg cyclophospha \d Arachidonic acid + indomethacin Â± 10 mide followed by saline produced no greater accumulation of 128 Â± I"* +Arachidonic acid + NDGA 99 Â±8 133Â± \2" hydroxyproline than did two doses of saline. +Arachidonic acid + ABTLiver 95 Â±9Lung In Vitro Production of Alkylating Metabolites. Cyclophos " Data are expressed as nmol mechlorethamine equivalents/15 min/mg pro phamide, in the absence of microsomes, exhibited a trace in. * Significantly different from liver (P < 0.05). amount of alkylating activity (1.3 Â±0.1 nmol mechloreth- ' Significantly less than cyclophosphamide + NADPH (P < 0.05). amine equivalents/assay). This value was subtracted from all ''Significantly less than cyclophosphamide + arachidonic acid (P < 0.05). 545

Table 2 Index of efficiency for cvclophosphamide as a reducing cosuhsirate of pear to affect the antitumor activity of Cyclophosphamide (30- PPIIP 32). The index of efficiency is Ihe fraction of PPA produced. This unilless value Pulmonary toxicity due to Cyclophosphamide is well docu was calculated by dividing the amount of PPHP reduced to PPA by the total amount of PPA + PPHP. A value of 1.0 indicates complete reduction to PPA. mented in the literature, but the mechanisms responsible for Indomclhacin. NDGA. or ABT (200 >iM) was added to the incubation i min prior this organ-specific damage remain largely uncharacterized. Spe to the substrate. Data are expressed as mean Â±SE of determinations from three cifically, the role of MFO-mediated metabolism in the lung separate mierosomal preparations. injury produced by this drug, although widely accepted, has not Index of efficiency been firmly established. Acrolein is the metabolite of Cyclo microsomes microsomes phosphamide most often suggested as being responsible for CosubstrateBHA (4 mgprotein)0.482 (1 mgprotein)0.473 lung injury. Acrolein formation has been detected following the (lOO^M) Â±0.021Â° +IndomethacinCyclophosphamide 0.4580.3650.0630.028Â°0.449 Â±0.0820.323 infusion of Cyclophosphamide into isolated perfused rat lung Â±0.030Â° (17), but it is not clear what metabolic pathway is responsible alone (100 UM) or whether sufficient acrolein is produced to explain the ob + Indomethacin 0.387 0.054* 0.304 Â±0.023* +NDGA 0.373 0.018* 0.223 Â±0.015*-' served, and delayed, injury characteristic of Cyclophosphamide. +ABTMicrosomes 0.4000.1670.1700.1930.014Â°0.029*0.0060.0120.0300.2170.183Â±0.012*-'0.1 The NADPH-supported production of alkylating metabolites

alone+ 22Â±0.0060.143 is apparently much lower in mouse (Table 1) than in rat (17) Indomethacin+ Â±0.0070.130 lung microsomes. Species differences in metabolic activation NDGA+ABTLiver Â±0.0270.1 are, therefore, possible. It is also possible that the commonly 0.007Lung 43 Â±0.009 * Significantly different from microsomes alone (P < 0.05). accepted scheme for Cyclophosphamide metabolism is unrelated * Significantly different from inhibitor alone (P < 0.05). to both the therapeutic and toxic effects of this drug. c Significantly different from the corresponding liver system (P < 0.05). The incorporation of radiolabeled thymidine into pulmonary DNA, which is indirectly related to the extent of the initial lung mediate the production of any alkylating metabolites in the injury (18, 33), was maximal 7 days after treatment with Cyclo presence of either an NADPH-generating system or arachidonic phosphamide. If the inhibition of MFOs prevents the activation acid (data not shown). of Cyclophosphamide to toxic metabolites, then pretreatment Reducing Cosubstrate Potential of Cyclophosphamide. The with inhibitors of MFO activity should prevent the lung damage ability of Cyclophosphamide to act as a reducing Cosubstrate for due to this drug. Similarly, if a preliminary dose of Cyclophos PPHP is shown in Table 2. An index value of approximately phamide protects against a subsequent dose, then the pattern 0.5 in the presence of BHA required 4 mg liver mierosomal of thymidine incorporation seen with two doses of Cyclophos protein or 1 mg lung mierosomal protein, in accordance with phamide should be altered to reflect that protective effect. the known difference in the activity of PHS in these two tissues. With the exception of ABT, none of the MFO inhibitors Using these same protein concentrations, Cyclophosphamide employed in the current study altered the typical pattern of produced index values of 0.323 and 0.365 in lung and liver pulmonary thymidine incorporation seen after Cyclophospha microsomes, respectively. Compared to Cyclophosphamide mide. This may have been due to an absence of an effect on alone, index values were unaffected by preincubation of liver or MFO-mediated bioactivation of Cyclophosphamide by lung tis lung microsomes with indomethacin or NDGA, while ABT was sue, even though the doses used have been effective in altering able to significantly decrease the index value in lung, but not the activation of other lung toxins (18-20). ABT did not alter liver, microsomes. Microsomes from lung and liver exhibited a the usual peak of thymidine incorporation seen on day 7 after low index number in the absence of any cosubstrates or cofac- Cyclophosphamide, although it did prevent increases on days 3 tors, in accordance with previous data (15). The index values and 10. The reason for this is unclear from the present results. produced by the inhibitors in the absence of Cyclophosphamide In contrast, BI significantly enhanced thymidine incorporation, were not different from the values for microsomes alone. and both BI and PBX significantly increased lung hydroxypro- Heating of mierosomal fractions at 100Â°Cfor 10 min de line content after Cyclophosphamide. These results indicate that stroyed their ability to reduce PPHP to PPA. in contrast, the lung damage due to Cyclophosphamide is not dependent on cytosolic fraction from liver and lung (0.5 mg protein) exhibited MFO activation and that another enzyme system, or systems, a high index value (0.483 Â±0.015 and 0.477 Â±0.022, respec may be involved in the toxification process. tively), which was retained even following boiling. This may be Pretrcatment of mice with inhibitors of arachidonic acid attributable to various cytosolic compounds, such as cysteine, which are capable of reducing PPHP by nonenzymic means.5 metabolism attenuated the usual increases in thymidine incor poration seen after Cyclophosphamide. While the precise mech anism of this protective effect cannot be determined from the current study, one possible explanation may involve the cooxi- DISCUSSION dation of Cyclophosphamide by PHS. PHS can activate a num Cyclophosphamide is an orally effective alkylating oxazo- ber of compounds of diverse chemical structure (13, 34-37) and it is possible that the 4-hydroxylation step that activates Cyclo phosphorine used as an immunosuppressant in the treatment of a wide variety of malignant and inflammatory disorders (2, phosphamide can be accomplished through this pathway. Al 29). Cyclophosphamide itself is inactive and is thought to be ternately, an entirely novel pathway may be involved. Although activated by the cytochrome P-450 MFO system (9). Cyclo liver is relatively poor in PHS, lung is known to have a much phosphamide is clearly metabolized by the MFO system of greater activity of this enzyme complex ( 13). Inhibitors of M FO hepatic microsomes (10), and this metabolism leads eventually activity, such as BI and PBX, may serve to shift the site of to the alkylator phosphoramide mustard and acrolein. How activation from liver to lung, where PHS or other peroxidases ever, neither SKF 525A nor phÃ©nobarbitalpretreatments ap- are the predominant oxidative enzymes, thus explaining the increased toxicity seen with these agents. 5 L. J. Marnett. personal communication. Consistent with the results of the thymidine incorporation 546

Downloaded from cancerres.aacrjournals.org on October 5, 2021. © 1991 American Association for Cancer Research. COOXIDATION AND LUNG TOXICITY OF ( Y( I OI'HOSI'HAMIDE experiments, the MFO inhibitors, except ABT, did not prevent the conclusion that activation of cyclophosphamide occurs, at the accumulation of lung hydrox\ proline in cyclophosphamide- least partially, by different enzyme systems in lung and liver. treated mice. ABT is a suicide inhibitor of many P450 isozymes MFO inhibitors decreased alkylating activity in liver, but not and has been shown to be effective in lung (19. 20). However. lung, while inhibitors of PHS were effective in lung, but not ABT may inactivate many other heme enzymes, such as PHS. liver. The sole exception to this was ABT, which may inhibit explaining why it prevented the development of pulmonary multiple heme-dependent enzyme systems and reduced alkyl fibrosis due to cyclophosphamide and produced levels of hy- ating activity in microsomes in both tissues. None of the inhib droxyproline that were similar to those in animals pretreated itors totally abolished production of alkylating substances. This with indomethacin. NDGA, or aspirin. This possibility is sup may be due to inadequate doses or the presence of multiple ported by the ability of ABT to block the reduction of PPHP activation pathways, such that no single inhibitor could be in lung microsomes. maximally effective. This latter possibility has been postulated Additional explanations for the protective effect seen with with the lung toxin 3-methylindole (16). the inhibitors of arachidonic acid metabolism include blockade With few exceptions (15), compounds that are cooxidized by of the inflammatory response or alterations in the ability of PHS are efficient electron donors to the peroxidase portion of macrophages and neutrophils to produce the products of ara this enzyme. The ability of a xenobiotic to serve as a reducing chidonic acid metabolism which can influence fibrotic proc cosubstrate can be measured by following the reduction of a esses. Indomethacin and NDGA can prevent or reduce the synthetic hydroperoxide. The reduction of PPHP to PPA has development of pulmonary fibrosis due to bleomycin (21, 38, proven to be a simple and reliable technique to assess cooxida- 39). However, the short duration of treatment used in the tion potential (15, 27). BHA, a highly active reducing cosub current study would not be expected to have major antiinflam- strate (15), was used to determine the amount of liver and lung matory effects. microsomal protein needed to assess the cooxidation potential. Cyclophosphamide itself can have effects on arachidonic acid Cyclophosphamide proved to be a moderately good reducing metabolism and may activate pulmonary immunocompetent cosubstrate for peroxidase activities in both lung and liver cells to increase lung damage (2) or alter pulmonary glutathione microsomes when present at the same concentration as BHA. to decrease prostaglandin E; and increase thromboxane B2 However. 4 times more liver microsomal protein was required levels (40). Administration of prostaglandin synthase inhibitors to reach the same level of activity as found in lung. may block the inflammatory cascade involving prostanoids and PHS is irreversibly inactivated within the 3-min time period prevent lung damage due to initial cyclophosphamide-induced of the assay (15), and increasing concentrations of a substrate injury. Cyclophosphamide produces lipid peroxidation in lung tend to enhance the reduction of PPHP, perhaps as a conse microsomes and liver supernatant (12, 41) and may enhance its quence of more reduction before enzyme inactivation. Since own toxicity by providing additional membrane lipids that are only one concentration of cyclophosphamide was used in the then shunted into the arachidonic acid pathway, thereby in current study, it is possible that higher index values could be creasing the inflammatory response. obtained, although this would prevent useful comparisons to Protection against lung damage due to a high dose of cyclo other substrates. The inability of indomethacin to block the phosphamide by a previous nonlethal dose of the same drug reduction of PPHP is not surprising, since this compound has has been reported using ventilation rate as an indicator of lung no effect on the peroxidase portion of PHS. NDGA is an injury. A protective effect could be related to inhibition of MFO antioxidant compound which should be an excellent reducing activity by low doses of this drug (42, 43). However, the results cosubstrate. Its lack of a significant effect on cyclophosphamide of the current study fail to support this conclusion. Using the cooxidation is currently unexplained, although it may be related same pretreatment regimen as Collis et al. (5), the pattern of to self-inactivation of PHS during the preincubation period. thymidine incorporation and total lung hydroxyproline content These data clearly show that cyclophosphamide can serve as was unaffected, when compared with a single dose of cyclo a cooxidation substrate and that in vivo treatment with PHS phosphamide. Other data indicated that two 100 mg/kg doses inhibitors can diminish the lung toxicity of this drug. Although of cyclophosphamide extended but did not greatly alter the the results of the current study do not prove a role for PHS in patterns of thymidine incorporation, compared to those seen the bioactivation of cyclophosphamide to lung-toxic metabo with a single dose. This suggests that additional lung cell lites, it is clear that such a pathway is a potential candidate for damage and resultant cell proliferation are only modestly af this process. Other peroxidase systems, such as myeloperoxi- fected by a "priming" dose. dase or lipoxygenases, might also be involved, although the The NBP assay has been validated as an accurate measure of absence of any alkylating metabolite production in lung or liver the production of alkylating metabolites from cyclophospha cytosol would argue against such possibilities. mide (44). The alkylating activity measured in liver microsomes Previous work on cyclophosphamide-induced lung injury has in the presence of NADPH was similar to that reported by implicated oxidative processes (12, 17). The current data do Hipkens et al. (44) in 10 strains of mice. Alkylating activity not conflict with these findings but do suggest a different source was also evident in lung microsomes but appeared to have for the toxic metabolites. These results are also consistent with differing cofactor requirements from that in liver. In lung the protective effects seen with certain antioxidant substances microsomes, greater levels of alkylating metabolites were gen used to treat the bladder injury associated with cyclophospha erated in the presence of arachidonic acid than with NADPH, mide and suggest that even more successful therapy might be while in liver microsomes the converse was true. However, obtained with various nonsteroidal antiinflammatory agents, quantitative comparisons of arachidonic acid- and NADPH- should the high levels of PHS found in the bladder be involved supported formation of alkylating metabolites is problematic in this clinically significant toxicity. due to the rapid (3 to 5 min) self-inactivation of PHS in vitro In summary, the lung cell damage and fibrosis that result (15). from a single dose of cyclophosphamide are reduced in severity The effects of inhibitors of MFO or PHS activity substantiate by inhibitors of arachidonic acid metabolism. Although com- 547

by prostaglandin H svnthase and mixed-function oxidases in goat lung and pÃeteprotection was not obtained with any of these drugs, only liver microsomes. Biochem. Pharmacol.. 37: 4359-4366. 1988. limited treatments were given and no attempt was made to 17. Patel, J. M. Metabolism and pulmonary toxicity of cyclophosphamide. optimize the dose or duration of therapy. In contrast, lung Pharmacol. Ther.. 47: 137-146. 1990. 18. Kehrer. J. P.. and Witschi. H. P. Effects of drug metabolism inhibitors on injury and fibrosis were not prevented by most MFO inhibitors butvlated hydroxytoluene-induced pulmonary toxicity in mice. Toxicol. Appi. or by a second dose of the drug. Together with additional Pharmacol., S3: 333-342, 1980. studies showing arachidonic acid-supported production of al- 19. Huijzer. J. C.. Adams, J. D., Jr., Jaw. J-Y.. and Yost. G. S. Inhibition of 3- nicthvlindolc bioactivation by the cylochrome P-450 suicide substrates I- kylating metabolites by lung microsomes in vitro, these data aminoben/otria/ole and u-niethylben/vlaminoben/otria/ole. Drug Metab. support the possibility that the PHS enzyme system is involved Dispos.. 17: 37-42. 1989. 20. Mathews. J. M.. Dostal. L. A., and Bend. J. R. lnacti\ation of rabbit in the lung toxicity of cyclophosphamide. These results high pulmonary cytochrome P-450 in microsomes and isolated perfused lung b\ light the need to study further the mechanisms of cyclophos- the suicide substrate l-aminobenzotriazole. J. Pharmacol. Fxp. Ther., 235: phamide-induced injury, with special consideration given to 186-190, 1985. 21. Thrall, R. S.. McCormick. J. R.. Jack. R. M.. McReynolds. R. A., and Ward. metabolic cooxidation, drug combinations, multiple doses of P. A. Bleomycin-induced pulmonar) lÃ¬brosisinIhe ral. Am. J. Pathol., 95: the same drug, and the time dependency of these interactions. 117-130. 1979. It will also be important to determine whether drugs such as 22. Kehrer. J. P. Collagen production rales following acule lung damage induced by butylated hydroxyloluene. Biochem. Pharmacol., 31: 2053-2058. 1982. aspirin or indomethacin affect the therapeutic activity of cyclo 23. Hance, A. J.. and Crystal, R. G. Collagen, in: R. G. Crystal (ed.). The phosphamide. Biochemical Basis of Pulmonary Function, pp. 215-271. New York: Marcel Dekker. Inc.. 1976. 24. Witschi. H. P.. Tryka. A. F.. and l.indenschmidt. R. L. The many faces of an increase in lung collagen. Fundam. Appi. Toxicol.. 5: 240-250. 1985. ACKNOWLEDGMENTS 25. Il/haki. R. F.. and Gill. D. M. A micro-biuret method for estimating proteins. Anal. Biochem.. 9: 401-410. 1964. The authors would like to thank Dr. Pat Randall for his advice and 26. Sladck. N. E. Metabolism of cyclophosphamide by rat hepatic microsomes. suggestions concerning the statistical analysis for this manuscript and Cancer Res., 31: 901-908. 1971. Thomas Devor for his expert technical assistance. 27. Weller. P. E., Markey, C. M.. and Marnett. L. J. En/ymatic reduction of 5- phenyl-4-pentenyl-hydroperoxide: detection of peroxidases and identification of peroxidase reducing substrates. Arch. Biochem. Biophys.. 243: 633-643. REFERENCES 1985. 28. Zivin. J. A., and Bartko. J. J. Statistics for disinterested scientists. Life Sci., 1. Kehrer. J. I'., and Kacew, S. Systemically applied chemicals thai damage 18: 15-26, 1976. 29. Brock, N. Oxazaphosphorine cylostalics: past, present and future. Cancer lung tissue. Toxicology, 35: 251-293. 1985. Res., 49: 1-7, 1989. 2. Cooper. J. A. D.. Jr.. While. D. A., and Matthay. R. A. Drug-induced pulmonary disease. Am. Rev. Respir. Dis.. 133: 321-340. 1986. 30. Hart, L. G., and Adamson. R. H. Effect of microsomal en/yme modifiers on toxicity and therapeutic activity of cyclophosphamide in mice. Arch. Int. 3. Weiss. R. B.. and Muggia. F. M. Cyloloxic drug-induced pulmonary disease: Pharmacodyn.. 180: 391-401. 1969. update 1980. Am. J. Med.. 68: 259-268. 1980. 31. Hill. D. L.. Laster. W. R., and Struck. R. F. Enzymatic metabolism of 4. Gould. V. E.. and Miller. J. Sclerosing alveolilis induced by cyclophospha mide. Am. J. Palhol.. */: 513-520. 1975. cyclophosphamide and nicotine: production of a toxic cyclophosphamide 5. Collis. C".H., Wilson, C. M.. and Jones. J. M. Cyclophosphamidc-induced metabolite. Cancer Res.. 32: 658-665. 1972. 32. Sladek. N. E. Therapeutic efficacy of cyclophosphamide as a function of lung damage in mice: protection by a small preliminary dose. Br. J. Cancer. inhibition of its metabolism. Cancer (Phila.). 32: 1848-1854, 1972. 41: 901-907, 1980. 33. Martin. F. M.. and Witschi, H. P. Cadmium-induced lung injury: cell kinetics 6. Kumar, R. K., Truscoll. J. Y.. Rhodes, G. C., and Lykke. A. W. J. Type 2 and long-term effects. Toxicol. Appi. Pharmacol.. 81):215-227, 1985. pneumocvte responses lo cyclophosphamide-induced pulmonary injury: func 34. Lau. S. S., and Monks, T. J. Co-oxidation of 2-bromohydroquinone b> renal tional and morphological correlation. Br. J. Exp. Palhol.. 69: 69-80. 1988. prostaglandin synthase. Drug Mctab. Dispos.. 15: 801-807, 1987. 7. HÃ¤kkinen.P. J.. Morse. C. C.. Martin. F. M.. Dalbey. W. E.. Haschek. W. 35. Marnett, L. J.. and Eling. T. E. Cooxidation during prostaglandin biosyn M., and WÂ¡ischi.H. R. Potentiating effects of oxygen in lungs damaged by thesis: a pathway for the metabolic activation of xenobiotics. Rev. Biochem. methylcvclopcntadienyl manganese tricarbonyl. cadmium chloride, oleic acid Toxicol.. 5: 135-172. 1983. and antilumor drugs. Toxicol. Appi. Pharmacol.. 67: 55-69. 1983. 36. Sivarajah. K.. Laster, J. M.. and Eling, T. M. Prostaglandin synthctasc- 8. Witschi. 11.. Godfrey. G.. Frome, E., and Lindenschmidt. R. C. Pulmonary dependent cooxidation of (Â±)-bcnzo(a)pyrene-7,8-dihydrodiol by human lung toxicity of eytostatic drugs: cell kinetics. Fundam. Appi. Toxicol.. fi: 253- and other mammalian tissues. Cancer Res.. 41: 1834-1839, 1981. 262. 1987. 37. Thompson. D. D.. Cha, Y-N.. and Trush, M. A. The peroxidase-dependent 9. Alarcon, R. A.. Sludies on the in viro formation of acrolein: 3-hydroxypropyl- activation of butylated hydroxyanisole and butylated hydroxytoluene (Bill) Cancermercapturic Treat. acid Rep.. as an60: index 327-335." of cyclophosphamide 1976. (NSC-26271 ) activation. to reactive intermediates. J. Biol. Chem.. 264: 3957-3965. 1989. 38. Chandler. D. B.. and Young. K. The effect of diclofenac acid (Voltaren) on 10. Cohen. J. L.. and Jao. J. Y. En7.ymatic basis of cyclophosphamide activation blcomycin-induced pulmonary fibrosis in hamsters. Prostaglandins Leuko- by hepatic microsomes of the rat. J. Pharmacol. Exp. Ther.. 174: 206-210. trienes Essential Fatty Acids. 3/i: 9-14. 1989. 1970. 39. Phan. S. H.. and Kunkel. S. L. Inhibition of blcomycin-induced pulmonary 11. Brock. N., and Hohorst, H. J. Metabolism of cyclophosphamide. Cancer fibrosis by nordihydroguiaretic acid. Am. J. Pathol., 124: 343-352. 1986. (Phila.). 21):900-904. 1967. 40. Cooper. J. A. D.. Jr.. and Merrill. W. W. Modulation of endoperoxide 12. Palei. J. W. Stimulation of cyclophosphamide-induced pulmonary microso- product levels and cyclophosphamide-induced injury by glutathione repletion. mal lipid peroxidation by oxygen. Toxicology, 45: 79-91, 1987. J. Appi. Physiol.. 67: 2316-2322. 1989. 13. Eling, T. E., Thompson. D. C.. Foureman. G. L.. Curtis. J. F.. and Hughs. 41. Berrigan. M. J.. Struck. R. F.. and Gurtoo. H. L. Lipid peroxidation by M. F. Prostaglandin H synlhase and xenobiolic oxidation. Annu. Rev. cyclophosphamide. Cancer Biochem. Biophys.. 9: 265-270. 1987. Pharmacol. Toxicol.. 30: 1-45. 1990. 42. Gurtoo. H. L.. Bansal. S. K., Pavelic, Z., and Struck. R. F. Effects of the 14. Marnett. L. J.. Wlodawer. P.. and Samuelsson. B. Co-oxygenation of organic induction of hepatic microsomal metabolism on the toxicity of cyclophos substrates by the prostaglandin synthetase of sheep vesicular gland. J. Biol. phamide. Br. J. Cancer. 51: 67-75. 1985. Chem.. 250: 8510-8517. 1975. 43. Patel. J. W., and Block. E. R. Cyclophosphamide-induced depression of the 15. Markey. C. M.. Alward. A.. Weller, P. E.. and Marnett. L. J. Quantitative antioxidant defense mechanisms of Ihe lung. Exp. Lung Res., 8: 153-165. studies of hydroperoxide reduction by prostaglandin H synthase. Reducing 1984. substrate specificity and the relationship of peroxidase to cyclooxygenase 44. Hipkens, J. H.. Struck, R. F., and Gurtoo. H. L. Role of aldehyde dehydrog- activities. J. Biol. Chem.. 262: 6266-6279. 1987. enase in the metabolism-dependent biological activity of cyclophosphamide. 16. Formosa. P. J.. and Bray. T. M. Evidence for metabolism of 3-methylindole Cancer Res.. 41: 3571-3583. 1981.

548

Downloaded from cancerres.aacrjournals.org on October 5, 2021. © 1991 American Association for Cancer Research. Cooxidation of Cyclophosphamide as an Alternative Pathway for Its Bioactivation and Lung Toxicity

Robin D. Smith and James P. Kehrer

Cancer Res 1991;51:542-548.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/51/2/542

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/51/2/542. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.