- Email: [email protected]
Preferential kill of hypoxic EMT6 mammary tumor cells by the bioreductive alkylating agent porfiromycin
Advan.
Pergamon
Regul., Vol. 35, pp. 117-130, 1995 Copyright 0 1995 Elsevier Science Ltd Printed in Great Britain. All rights reserved Ol%S-2571/951$29.(Xl Enzyme
00652571(94)00015-S
PREFERENTIAL KILL OF HYPOXIC EMT6 MAMMARY TUMOR CELLS THE BIOREDUCTIVE ALKYLATING AGENT PORFIROMYCIN
BY
ALAN C. SARTORELLI, MICHAEL F. BELCOURT, WILLIAM F. HODNICK, SUSAN R. KEYES*, CHRIS A. PRITSOS and SARA ROCKWELL Departments of Pharmacology and Therapeutic Radiology and Developmental Therapeutics Program, Yale Cancer Center, Yale University School of Medicine, New Haven, CT 06520 INTRODUCTION
Solid tumors of both humans and animals contain cells which are distal to the tumor vasculature and consequently exist in an environment of nutritional deficiency, low pH and hypoxia (l-4). Hypoxic cells constitute a therapeutically resistant component of solid tumors. Since oxygen sensitizes cells to the cytotoxic action of X-rays, these cellular elements are resistant to ionizing radiation (1). They are also relatively insensitive to many antineoplastic agents for a variety of reasons (5-12). These include the fact that (a) a number of cancer chemotherapeutic agents have difficulty diffusing into hypoxic regions of tumors in therapeutically adequate concentrations, (b) hypoxic cells progress slowly through the cell cycle or are blocked in their transition through the cycle and, therefore, are resistant to cycle active anticancer drugs, and (c) hypoxia produces amplification of genes capable of conferring resistance to a variety of agents. The oxygen deficiencies within the hypoxic regions of solid tumors lead to an environment conducive to reductive events, thereby creating an opportunity to preferentially attack hypoxic cells with agents that require enzymatic reduction to create electrophilic species. We and others have employed this concept, initially delineated in 1972 (13), to demonstrate that the mitomycin antibiotics preferentially kill hypoxic tumor cells relative to their aerobic counterparts (5-7, 14-18). The cytotoxicity of the mitomycin antibiotics requires bioactivation of these agents to *Present address: Biomeasure, Inc., 27 Maple Street, Mifford, MA 01757-3650. +Present address: Department of Nutrition, University of Nevada, Rem), NV 89557 117
118
A. C. SARTORELLI
et
al.
produce mono- and b&DNA adducts (19-21). A number of enzymes have been demonstrated to be capable of reductively activating the mitomycins, including NADPH:cytochrome P-450 oxidoreductase (EC 1.6.2.4) (22-30), NADH:cytochrome b, oxidoreductase (EC 1.6.2.2) (31, 32), NAD(P)H: quinone oxidoreductase (EC 1.6.99.2; DT-diaphorase) (33-38), xanthine: oxygen oxidoreductase (EC 1.2.3.2; xanthine oxidase) (24, 29, 39), and xanthine: NAD+ oxidoreductase (EC 1.2.1.37; xanthine dehydrogenase) (40, 41). Mitomycin C has been considered by our laboratories to be the prototype bioreductive alkylating agent. We have characterized this agent extensively, demonstrating in mice bearing established solid tumors that this antibiotic in combination with X-irradiation produced additive or greater-than-additive cytotoxicity to tumor tissue, without increased toxicity to normal tissues (for appropriate references see 14). The antineoplastic efficacy of this combination was considered to result from the preferential kill of aerobic tumor cells by the ionizing radiation and of hypoxic cells by mitomycin C. These findings led to the conduct of two prospective randomized clinical trials at Yale University, evaluating mitomycin C as an adjunct to radiation therapy in patients with head and neck cancer; these studies demonstrated significantly improved local control and therapeutic gain with the combination (42, 43). These results have led us to look for agents with greater differential kill of hypoxic and aerobic cells, which might be employed to minimize toxicity to oxygenated host tissues and to maximize the destruction of the hypoxic cells of solid tumors. Porfiromycin has been found to be such an agent, being essentially equivalent to mitomycin C on a molar basis in its toxicity to hypoxic cells in vitro, but being much less toxic than mitomycin C to aerobic cells (15-18,44,45). The present report summarizes data from our laboratories on the effects of porfiromycin on hypoxic and aerobic EMT6 murine mammary tumor cells. MATERIALS
AND
METHODS
Porfiromycin, [sH]porliromycin and mitomycin C were gifts from Dr Terrance Doyle of Bristol Laboratories (Wallingford, CT). NADPH:cytochrome P-450 oxidoreductase was purified from phenobarbital-treated rabbit liver as described previously (28). All other materials were purchased from commercial sources: gases (Presto Welding Service Center, North Haven, CT); fetal bovine serum and Waymouth’s medium (Grand Island Biological Co., Grand Island, NY); and NADH and NADPH (Sigma Chemical Co., St Louis, MO). All experiments were performed using EMT6 mouse mammary tumor cells, which were maintained by alternate passage in vivo and in vitro as
HYPOXIC
CELL ACTIVITY
OF PORFIROMYCIN
119
detailed elsewhere (46). Cell culture studies were performed in Waymouth’s medium supplemented with 15% fetal bovine serum and antibiotics, using exponentially growing monolayers in glass milk dilution bottles. Hypoxia was produced by gassing cultures with a humidified mixture of 95% N,--5% CO, for 2 hr. Mitomycin C and porfiromycin were dissolved in 70% ethanol and small vols of drug were added to the hypoxic cultures without compromising the hypoxia, as described previously (47). The viability of the cells was assayed by measuring the ability of individual cells to form colonies (46). Each experiment included controls subjected to all experimental manipulations and incubated with the vehicle used to dissolve the drug under appropriate aerobic and hypoxic conditions. In vivo studies were performed using EMT6 tumors in 2.5- to 3-monthold BALB/cRw mice by previously published procedures (46,48). Tumors were implanted by the injection of 2 x 105 cells into the skin of the flank and were allowed to grow for 2 weeks, producing tumors approximately 100 cu mM in vol. Porfiromycin was administered i.p. as a single dose 1 hr before sacrifice. Mice were whole body irradiated, without anesthesia, with 15 Gy of 250 kV X-rays at a dose rate of approximately 2 Gy/min (17, 49). Radiation was begun 50 min after the drug treatment. Previous studies demonstrated that the toxicities of mitomycin C and radiation ipr vitro were independent and did not depend upon the sequence or timing of the mitomycin C and radiation treatments (50) and that intervals of 5 min to 24 hr between mitomycin C and X-rays in vivo gave similar cytotoxicities (49). Tumors were excised immediately after irradiation or 1 hr after porfiromycin treatment. Tumor cells were suspended by a combination of mechanical and enzymatic disruption, collected by centrifugation, counted and assayed for viability by determining their ability to form colonies in vitro (46, 48, 49). Uptake and efflux of porhromycin were measured using single-cell suspensions in Waymouth’s medium supplemented with 15’% fetal bovine serum (51). These suspensions were prepared from cells which had been grown in monolayer culture and allowed to recover from trypsinization for 2 hr at 37°C in an atmosphere of 95% air-5% CO,. Five-milliliter aliquots containing 5 x 105 cells/ml were placed in glass scintillation vials and gassed through rubber serum stoppers with 95% N,-5% CO, or 95% air-5% CO, for 2 hr at 37°C. [3H]Porfiromycin was added without compromising the hypoxia. After various time intervals, 0.3 ml aliquots were removed and cell-associated radioactivity was measured by centrifugation through oil (sihcone:mineral oil, 84:16, v/v) into 0.5 N perchloric acid. Radiolabel in the cell pellet was ascertained by scintillation spectrometry in Ultrafluor (National Diagnostics, Somerville, NJ). At the end of the 2 hr uptake period, cells were washed twice with warm medium
120
A. C. SARTORELLI
er al.
equilibrated with 95% air-5% CO, and resuspended in medium to the original concentration of 5 x 10s cells/ml. Cell-associated [sH]porfiromycin was then measured as a function of time after resuspension, using the techniques described above. Throughout the entire procedure, cells were shaken to prevent reattachment. For technical reasons, uptake and efflux studies were performed using suspension cultures (51). The cytotoxicity of porfiromycin was similar for EMT6 cells in monolayer and suspension cultures. Calculation of the intracellular concentration of portiromycin used the following parameters: 5.45 nmol = 1.78 a 0.14 X 106 cpm and cell vol = 2.2 pl. Cell sonicates were prepared from exponentially growing monolayer cultures by detaching cells from flasks using 0.05% trypsin in PBS, washing with cold PBS and, subsequently, resuspending them in the appropriate buffer system. The cells in an ice bath were disrupted by three 5-set bursts with a Branson Model 160 sonicator set at 25% of maximum intensity. Cellular disruption was verified microscopically. Oxygen consumption was used as a measure of oxygen radical generation (52) and was monitored polarographically with a Gilson oxygraph equipped with a Clark electrode (53). The enzymatic activation of drug-induced oxygen consumption was assayed by resuspending EMT6 cells in cold PBS at a concentration of 3-5 x 106 cells/ml and disrupting them by sonication as described above. Final reaction mixtures consisted of 2 ml of EMT6 cell sonicate, 50 mM Tris-HCl buffer (pH 7.8), 1 mM diethylenetriaminepentaacetic acid, 0.5 mM NADPH, 0.5 mM NADH and drug or ethanol as a control. Drugs were dissolved in 70% ethanol to a final concentration of either 1 or 5 mM; dilutions of these stock solutions resulted in no greater than a 1% final concentration of ethanol in incubation flasks, a level that did not affect oxygen consumption. Assays were conducted at 37°C. Initially, cell sonicates and substrates were added to the respiratory cell and oxygen consumption was monitored to obtain a baseline rate for cell sonicate respiration. Once a baseline rate was established, either drug (in ethanol, O-l% final concentration) or ethanol (as a vehicle control) was added and the new rate of oxygen consumption was ascertained. The drug-induced rate of oxygen consumption was obtained by subtracting the rate prior to drug addition from the rate following its addition. This method of determining drug-induced oxygen consumption was used for all of the enzyme assays. Drug-induced oxygen consumption in the presence of NADPH-cytochrome P-450 oxidoreductase was conducted by incubation in 2 ml of 50 mM phosphate buffer (pH 7.4), 0.5 mM NADPH, 0.18 unit of NADPHcytochrome P-450 oxidoreductase and either drug or vehicle; assays were conducted at 30°C.
HYPOXIC
CELL ACTIVITY
OF PORFIROMYCIN
121
The detection of superoxide and hydroxyl radical generated as a consequence of drug activation in sonicates of EMT6 cells was performed using the HPLC procedure described previously (54). This qualitative assay is useful in identifying the oxygen radical species generated by biological systems and can be used to compare relative radical production by various agents in the same system. Prior to sonication, EMT6 cells were suspended at a concentration of 1.6-2.0 X 107 cells/ml in 50 mM Tris-HCI buffer (pi-l 7.8) with 1 mM diethylenetriaminepentaacetic acid. The reaction mixture consisted of 200 ~1 of cell sonicate, 100 mM 5,5-dimethyl-1-pyrroline N-oxide, 0.5 mM NADH, 0.5 mM NADPH and 4.0% ethanol or drug (in ethanol) in a final vol of 250 ~1. The reaction mixture was incubated for 6 min at 37”C, at which time 25 ~1 of the solution were injected into the HPLC system, which consisted of an Altex ultrasphere 3 p,M octadecylsilane 7.5 x 4.6 mm inside diameter column attached to a Bioanalytical Systems LC4 A electrochemical detector (Bioanalytical Systems, Inc., West Lafayette, IN). The detector potential was set at +0.6 V versus an Ag+.-AgCl reference electrode. A Rainin C-18 Microsorb Short-Ones column (Rainin Instrument Co., Woburn, MA) served as a guard column. The mobile phase consisted of 0.03 M citric acid (monohydrate), 0.05 M anhydrous sodium acetate. 0.05 M sodium hydroxide and 8.5% acetonitrile, brought to pH 5.1 with glacial acetic acid; the flow rate was 1.0 ml/min. The relative peak areas of the radicals generated by each drug were determined by cutting the appropriate peaks from the HPLC tracings and weighing them on an analytical balance. RESULTS
AND
DISCUSSION
The toxicities of mitomycin C and porfiromycin (Fig. 1) to hypoxic and aerobic EMT6 cells are shown in Figure 2. On a molar basis, the toxicities of porfiromycin and mitomycin C to EMT6 cells were essentially the same under hypoxia. In contrast, although both mitomycin C and porfiromycin were less toxic to cells under aerobic conditions, porfiromycin was considerably less cytotoxic than mitomycin C in air. Thus, even though the redox potentials of these two mitomycin antibiotics are the same, the addition of a methyl group onto the aziridyl nitrogen (Nla) of mitomycin C markedly decreases its aerobic cytotoxicity, without altering its toxicity under hypoxia. The relatively large difference in the sensitivities of hypoxic and oxgenated EMT6 cells to porfiromycin correlated with differences in the uptake of this antibiotic under these different conditions of oxygenation. Under both aerobic and hypoxic conditions, the rates of uptake of porfiromycin increased with increasing concentrations of radiolabeled drug (Table 1). Uptake consisted of two phases: a rapid phase, which
122
A. C. SARTORELLI
et al.
0 ,4
NH2
**II
3
N
a3 )k!sii
0
Mitomycin C
Porfllycin
FIG. 1. Structures of mitomycin C and porfiromycin.
occurred within 2 min, and a relatively slow phase, which occurred at longer times (51). Uptake occurring during the rapid phase was inhibited by approximately 25-30% when the experiments were performed at 0°C whereas the slower uptake was completely abolished at 0°C. The rapidly associating porfrromycin was blocked approximately 25% by excess unlabeled porfiromycin. The oxygenation state of the cells did not influence the rapid uptake of porfiromycin. These findings suggest that the rapid uptake of porfiromycin primarily represents non-specific binding of the
1
g
10-l
g
.
M .9 .z t
10
-2
I 10-J v) IO4 10-S 0.01
0.1
1 10 Concentration (p&d)
100
FIG. 2. Survival of EMT6 mouse mammary tumor cells exposed to different concentrations of mitomycin C or porlkomycin for 1 hr under hypoxic or aerobic conditions in vitro. Points are geometric means of 240 independent determinations of surviving fractions. The SEMs are shown where n 3 3 and where the error is larger than the point size. Key: Mitomycin C-hypoxic (A), mitomycin C-aerobic (A), porfiromycin-hypoxic (0), porfiromycin-aerobic (0). Reproduced from Ref. (55).
HYPOXIC
CELL ACIWITY
OF PORFIROMYCIN
123
TABLE 1. RATES OF PORFIROMYCIN UPTAKE INTO EMT6 CELLS IN AIR AND IN HYPOXIA
Condition Air Hypoxia
Concentration (PM) 3 10 20 0.2 3
Rate (cpmlhril.5 X 105 cells) 120 300 1100 x0 1000
Reproduced from Ref. (51)
drug. To achieve comparable rates of uptake during the slow phase, a lo-fold lower concentration of porfiromycin was needed in hypoxia than was required in air (Table 1). This difference in drug concentration approximated the difference in porfiromycin levels needed to produce equivalent cytotoxicities under aerobic and hypoxic conditions (51). To determine whether porfiromycin possesses the appropriate physicochemical properties to diffuse through the tumor mass and reach hypoxic cells in concentrations adequate to produce cytotoxicity, the response of solid EMT6 tumors in viva to this antibiotic was measured and the results are shown in Figure 3. The survival curve for tumor cells
Porfkomycin dose (&g) FIG. 3. Effect of varying doses of porfiromycin alone (0) or in combination with 1.5 Gy of X-irradiation (0) on the survival of cells in solid EMT6 tumors. Points: Means + SEMs; 3-8 determinations. Stippled area: Area of additive cytotoxicities calculated from the complete single-agent dose-response curves. Reproduced from Ref. (17).
124
A. C. SARTORELLI
et al.
explanted from mice treated with porfiromycin alone had a very shallow slope. This reflects the fact that approximately 80% of the viable cells in EMT6 tumors are aerobic (56); the overall response of these tumors to porfrromycin is therefore dominated by the response of the aerobic tumor cells. Aerobic tumor cells in vim, like those in vitro (Fig. 2), were relatively resistant to porfiromycin. To examine the effects of porfiromycin on hypoxic tumor cells in vim, we selectively depleted the aerobic tumor cell population by treating the tumors with a large dose of X-rays (15 Gy). Because hypoxic cells are exceedingly resistant to radiation, while aerobic cells are radiosensitive (48, 56), this treatment reduces the survival of the aerated tumor cell population to approximately O.Ol%, while approximately 20% of the hypoxic tumor cells survive (56). Therefore, the tumor cell population surviving irradiation is composed almost entirely of hypoxic cells. The effects of porfiromycin on hypoxic (radioresistant) tumor cells were greater than the effects on the overall tumor cell population. Isobologram analyses, performed using complete dose-response curves for porfiromycin alone and for radiation alone (14, 17,49), indicated that the combinations of 15 Gy of X-rays with 6 ug or 20 pg of porfiromycin per g of body weight produced surviving fractions that lay outside the area of additivity and were compatible with supra-additive effects. This probably reflects the fact that each agent is effective against a tumor cell population which is resistant to the other agent. These findings have led to a Phase I clinical trial of porfiromycin in combination with radiation therapy in patients with advanced head and neck cancer (57). The lower toxicity of porfiromycin to oxygenated cells permitted the use of a porfiromycin dose that was more than 2.5 times the dose used for mitomycin C, with no significant increase in drug toxicity (42, 43, 57). Of 21 patients treated in this series, ten remain alive with no evidence of disease, with a median follow-up of 18.5 months. These results have led to an ongoing Phase III trial at Yale University in patients with squamous cell carcinoma of the head and neck region, comparing long-term local control in patients receiving radiation therapy plus porfiromycin to patients receiving radiation therapy plus mitomycin C. The major difference in the cytotoxicities of porfiromycin to hypoxic and oxygenated EMT6 cells has prompted an evaluation of the importance of the metabolism of the antibiotic in producing this difference in sensitivity. To gain insights into the metabolic processes underlining the hypoxic/aerobic differential, the metabolism of mitomycin C and porfiromycin was compared under aerobic conditions, with the objective being to ascertain the reason for the difference between the aerobic cytotoxicities of these two agents. To ascertain whether mitomycin C and porfiromycin were capable of being activated to a state that generates oxygen radicals in a complete tumor cell system, drug-induced oxygen consumption was measured in EMT6
HYPOXIC
CELL ACTIVITY
OF PORFIROMYCIN
125
cell sonicates supplemented with sufficient NADH and NADPH to ensure that these substrates were not rate-limiting. Both of the compounds tested produced linear concentration-dependent rates of oxygen consumption in the presence of EMT4 cell sonicates and substrate (Fig. 4). The rate of oxygen consumption with mitomycin C was approximately twice that with porfirom ycin . The quantities of superoxide and hydroxyl radical generated by mitomycin C and porfiromycin in sonicates of EMT6 cells were measured and are compared in Table 2. The amounts of superoxide and hydroxyl radicals produced in the presence of mitomycin C were signifkantly greater than those produced by control sonicates. In contrast, radical production in the presence of porfiromycin did not differ from control values. These results presumably are a reflection of the greater capacity of one or more of the enzyme systems present in EMT6 cells to reductively activate mitomycin C, relative to porfiromycin, in air. To ascertain whether NADPH:cytochrome P-450 oxidoreductase was involved in the difference in oxygen consumption between mitomycin C and porfiromycin in sonicates of EMT6 cells, the rate of activated drug-induced oxygen consumption produced by this enzyme was measured and found to be identical for mitomycin C and porfiromycin (Fig. 5). The finding of no difference in the capacity of NADPHcytochrome P-450 oxidoreductase to reductively activate mitomycin C and portiromycin
FIG. 4. Mitomycin antibiotic-induced stimulation of oxygen consumption catalyzed by EMT6 cell sonicates. Reaction mixtures consisted of 2 ml of EMT6 celf aonicates (3-5 x 106 c&s/ml in 50 rnM Tris-HCl buffer, pH 7.8). 1 mM d~e~~~~~~~~c acid, 0.5 mu NADPH and 0.5 mM NADH. Drugs were added at the in#c&?d fintll conooatrations and the reactions were run at 37°C as previously described. 0, mitomycin C; a, portiromycin. Points: Average of at least four determinations: bars, SD. Reproduced from Ref. (25).
126
A. C. SARTORELLI
et al.
TABLE 2. RELATIVE QUANTITIES OF DRUG-INDUCED OXYGEN RADICALS FORMED IN EMT6 CELL SONICATES Compound
Relative production Superoxide Hydroxyl radical
Mitomycin C Porfiromycin
1.29 + 0.03* 0.97 2 0.20
1.18 zk 0.10’ 1.06 + 0.10
*Values with a significant difference of at least P < 0.05 (by the paired Student’s r test) when compared to control samples equated to 1.0. Each value represents the mean + SD from at least 5 determinations. Reproduced from Ref.(25)
makes NADH: cytochrome b, oxidoreductase, DT-diaphorase, xanthine oxidase and/or xanthine dehydrogenase candidate enzymes to explain the difference in the activation of these drugs under aerobic conditions. SUMMARY
Hypoxic cells in solid tumors represent a therapeutically resistant population that limits the curability of many solid tumors by irradiation and by most chemotherapeutic agents. The oxygen deficit, however, creates an environment conducive to reductive processes; this results in a major exploitable difference between normal and neoplastic tissues.
FIG. 5. Mitomycin antibiotic-induced stimulation of oxygen consumption catalyzed by NADPH-cytochrome P-450 oxidoreductase. Reaction mixtures consisted of 2 ml of 50 mM phosphate buffer (pH 7.4), 0.18 unit of NADPH-cytochrome c reductase, 1 mM diethylenetriaminepentaacetic acid and 0.5 mM NADPH. Drugs were added at the indicated final concentrations and the reactions were run at WC as previously described. 0, mitomycin C; 0, portIromycin. Points: average of at least four determinations; bars. SD. Reproduced from Ref. (25).
HYPOXIC CELL ACTIVITY
OF PORFIROMYCIN
127
The mitomycin antibiotics can be reductively activated by a number of oxidoreductases, in a process required for the production of their therapeutic effects. Preferential activation of these drugs under hypoxia and greater toxicity to oxygen-deficient cells than to their oxygenated counterparts are obtained in most instances. The demonstration that mitomycin C and porfiromycin, used to kill the hypoxic fraction, in combination with irradiation, to eradicate the oxygenated portion of the tumor, produced enhanced cytodestructive effects on solid tumors in animals has led to the clinical evaluation of the mitomycins in combination with radiation therapy in patients with head and neck cancer. The findings from these clinical trials have demonstrated the value of directing a concerted therapeutic attack on the hypoxic fraction of solid tumors as an approach toward enhancing the curability of localized neoplasms by irradiation.
ACKNOWLEDGEMENTS
This research was supported in part by Grants DHP-102B (A.C.S.) and EDT-62N (S.R.) from the American Cancer Society.
REFERENCES 1. J. E. MOULDER and S. ROCKWELL, Tumor hypoxia: its impact in cancer therapy, Cancer Metastasis Rev. 5, 313-341 (1987). 2. P. VAUPEL, F. KALLINOWSKI and P. OKUNIEFF, Blood Bow, oxygen, and nutrient suppIy, and metabolic microenvironment of human tumors: a review, Cancer Res. 49, 6449-6465 (1989). .3. J. E. MOULDER, J. DUTREIX, S. ROCKWELL and D. W. SIEMANN, Applicability of animal tumor data to cancer therapy in humans, Inr. J. Radiut. Oncol. Biol. Phys. 14, 913-927 (1988). 4. H. D. SOSTMAN, H. C. CHARLES, S. ROCKWELL, K. LEOPOLD, C. BEAM, D. MADWED, M. DEWHIRST, G. CONFER, D. MOORE, R. BURN and J. OLESON, Soft-tissue sarcomas: detection of metabolic heterogeneity with P-31 MR spectroscopy, Radiology 176, 837-843 (1990). 5. K. A. KENNEDY, B. A. TEICHER, S. ROCKWELL and A. C. SARTORELLI, The hypoxic tumor cell: a target for selective cancer chemotherapy, Biochem. Phnrmacol. 29, l-8 (1980). 6. S. ROCKWELL, Hypoxic cells as targets for cancer chemotherapy, pp. 157-172 in Development of Target-oriented Anticancer Drugs (Y.-C. CHENG, B. GOZ and M. MINKOFF, eds.) Raven Press, New York (1983). 7. A. C. SARTORELLI, Therapeutic attack of hypoxic cells of solid tumors: presidential address, Cancer Res. 48,775-778 (1988). Classification of 8. B. A. TEICHER, J. S. LAZO and A. C. SARTORELLl. antineoplastic agents by their selective toxicities toward oxygenated and hypoxic tumor ceils, Cancer Res. 41, 73-81 (1981). 9. I. F. TANNOCK, The relation between cell proliferation and the vascular system in a transplanted mouse mammary tumour, Brif. J. Cancer 22, 258-273 (I!%@. 10. L. SIMPSON-HERREN and P. E. NOKER, Diversity of penetratton of anticancer agents into solid tumors, Cell Proliferation 24, 355-365 (1991).
128
A. C. SARTORELLI
et al.
11. G. C. RICE, C. HOY and R. T. SCHIMKE, Transient hypoxia enhances the frequency of dihydrofolate reductase gene amplification in Chinese hamster ovary cells, Proc. Natl. Acad. Sci. USA 83, 5978-5982 (1986). 12. G. C. RICE, V. LING and R. T. SCHIMKE, Frequencies of independent and simultaneous selection of Chinese hamster cells for methotrexate and doxorubicin (adriamycin) resistance, Proc. Natl. Acad. Sci. USA 84, 9261-9264 (1987). 13. A. J. LIN, L. A. COSBY, C. W. SHANSKY and A. C. SARTORELLI, Potential bioreductive alkylating agents. I. Benzoquinone derivatives, J. Med. Chem. 15, 1247-1252 (1972). 14. S. ROCKWELL and A. C. SARTORELLI, Mitomycin C and radiation, pp. 125-139 in Interactions Between Antitumor Drugs and Radiation (B. T. HILL and A. S. BELLAMY, eds.) CRC Press, New York (1990). 15. S. R. KEYES, S. ROCKWELL and A. C. SARTORELLI, Porfnomycin as a bioreductive alkylating agent with selective toxicity to hypoxic EMT6 tumor cells in vivo and in vitro, Cancer Res. 45, 3642-3645 (1985). 16. P. M. FRACASSO and A. C. SARTORELLI, Cytotoxicity and DNA lesions produced by mitomycin C and porfiromycin in hypoxic and aerobic EMT6 and Chinese hamster ovary cells, Cancer Res. 46, 3939-3944 (1986). 17. S. ROCKWELL, S. R. KEYES and A. C. SARTORELLI, Pre-clinical studies of porfiromycin as an adjunct to radiotherapy, Radiat. Res. 116, 100-113 (1988). 18. R. S. MARSHALL and A. M. RAUTH, Oxygen and exposure kinetics as factors influencing the cytotoxicity of porfiromycin, a mitomycin C analogue, in Chinese hamster ovary cells, Cancer Res. 48, 5655-5659 (1988). 19. S.-S. PAN, Porfiromycin disposition in oxygen-modulated P388cells, Cancer Chemother. Pharmacol. 27, 187-193 (1990). 20. M. TOMASZ, C. S. HUGHES, D. CHOWDARY, S. R. KEYES, R. LIPMAN, A. C. SARTORELLI and S. ROCKWELL, Isolation, identification, and assay of [3H]-porfiromycin adducts of EMT6 mouse mammary tumor cell DNA: effects of hypoxia and dicumarol on adduct patterns, Cancer Commun. 3, 213-223 (1991). 21. R. BIZANEK, D. CHOWDARY, H. ARAI, M. KASAI, C. S. HUGHES, A. C. SARTORELLI, S. ROCKWELL and M. TOMASZ, Adducts of mitomycin C and DNA in EMT6 mouse mammary tumor cells: effects of hypoxia and dicumarol on adduct patterns, Cancer Res. 53. 5127-5134 (1993). 22. K. KENNEDY, S. G. SLIGAR; L. POLOMSKI and A. C. SARTORELLI, Metabolic activation of mitomycin C by liver microsomes and nuclei, Biochem. Pharmacol. 11, 2011-2016 (1982). 23. T. KOMIYAMA, T. OK1 and T. INUI, Activation of mitomycin C and quinone drug metabolism by NADPH-cytochrome P-450 reductase, J. Pharm. Dyn. 2, 407410 (1979). 24. N. R. BACHUR, S. L. GORDON, M. V. GEE and H. KON, NADPH cytochrome P-450 reductase activation of quinone anticancer agents to free radicals, Proc. Natl. Acad. Sci. USA 76, 954-957 (1979). 25. C. A. PRITSOS and A. C. SARTORELLI, Generation of reactive oxygen radicals through bioactivation of mitomycin antibiotics, Cancer Res. 46, 3528-3532 (1986). 26. T. KOMIYAMA, T. KIKUCHI and Y. SUGIURA, Generation of hydroxyl radical by anticancer quinone drugs, carbazilquinone, mitomycin C, aclacinomycin A and adriamycin, in the presence of NADPH-cytochrome P-450 reductase, Biochem. Pharmacol. 31, 3651-3656 (1982). 27. S.-S. PAN, P. A. ANDREWS and C. J. GLOVER, Reductive activation of mitomycin C metabolites catalyzed by NADPH-cytochrome P-450 reductase and xanthine oxidase, J. Biol. Chem. 259, 959-966 (1984). 28. S. R. KEYES, P. M. FRACASSO, D. C. HEIMBROOK, S. ROCKWELL, S. G. SLIGAR and A. C. SARTORELLI, Role of NADPH:cytochrome C reductase and DT-diaphorase in the biotransformation of mitomycin C, Cancer Res. 44, 5638-5643 (1984). 29. S.-S. PAN, T. IRACKI and N. R. BACHUR, DNA alkylation by enzyme-activated mitomycin C, Mol. Pharmacol. 29, 622-628 (1986).
HYPOXIC CELL ACTIVITY
OF PORFIROMYCIN
129
30. H. F. J. BLIGH, A. BARTOSZEK, C. N. ROBSON, I. D. HICKSON, C. B. KASPER, J. D. BEGGS and C. R. WOLF, Activation of mitomvcin C bv NADPH: cytocbrome P-450 reductase, Cancer Res. 50, 7789-7792 (1990). _ . 31. W. F. HODNICK and A. C. SARTORELLI, Reductive activation of mitomycin C by NADH:cytochrome bs reductase, Cancer Res. 53,4907-4912 (1993). 32. J. FISHER and R. OLSEN, Mechanistic aspects of mitomycin C activation by Aavoprotein transhydrogenases, Develop. Biochem. 21, 240-243 (1982). 33. C. A. PRITSOS, L. L. PARDINI, A. J. ELLIOTT and R. S. PARDINI, Relationship between the antioxidant enzyme DT-diaphorase and tumor response to mitomycin C treatment, pp. 713-716 in Oxygen Radicals in Biology and Medicine (M. G. SIMIC, K. A. TAYLOR, J. F. WARD and C. VON SONNTAG. eds.) Plenum. New York (1989). 34. A. BEGLEITER, E. ROBOTHAM, G. LACEY and M. K. LEITH, Increased sensitivity of quinone resistant cells to mitomycin C. Cancer L,effers 49, 373-176 (1989). 35. A. BEGLEITER, E. ROBOTHAM and M. K LEITH, Role of NAD(P)H:(quinone acceptor) oxidoreductase (DT-diaphorase) in activation of mitomycin C under hypoxia, Mol. Pharmacol.
41, 677-682 (1992).
36. D. SIEGEL, N. W. GIBSON, P. C. PREUSCH and D. ROSS, Metabolism of mitomycin C by DT-diaphorase: role in mitomycin C-induced DNA damage and cytotoxicity in human colon carcinoma cells, Cancer Res. SO, 7483-7489 (1990). 37. D. SIEGEL, H. BEALL, C. SENEKOWITSCH, M. KASAI, H. ARAI, N. W. GIBSON and D. ROSS, Bioreductive activation of mitomycin C by DT-diaphorase, Biochemistry. 31, 7879-7885 (1992). 38. H. D. BEALL, R. T. MULCAHY, D. SIEGAL, R. D. TRAVER, N. W. GIBSON and D. ROSS, Metabolism of bioreductive antitumor compounds by purified rat and human DT-diaphorases, Cancer Res. 54, 3196-3201 (1994): 39. K. A. KENNEDY, E. G. MIMNAUGH, M. A. TRUSH and B. K. SINHA, Effects of glutathione and ethylxanthate on mitomycin C activation by isolated rat hepatic or EMT6 mouse mammary tumor nuclei, Cancer Res. 45, 4071-4076 (1985). 40. D. L. GUSTAFSON and C. A. PRITSOS, Bioactivation of mitomycin C by xanthine dehydrogenase from EMT6 mouse mammary carcinoma tumors, J. Nat/. Cuncer Insf. 84, 1180-1185 (1992). 41. D. L. GUSTAFSON and C. A. PRITSOS, Enhancement of xanthine dehydrogenase mediated mitomycin C metabolism by dicumarol, Cancer Res. 52, 6936-6939 (1992). 42. J. B. WEISSBERG, Y. H. SON, R. J. PAPAC, C. SASAKI, D. B. FISCHER, R. LAWRENCE, S. ROCKWELL, A. C. SARTORELLl and J. J. FISCHER, Randomized clinical trial of mitomycin C as an adjunct to radiotherapy in head and neck cancer. Int. J. Radial. Oncol. Biol. Phys. 17, 3-9 (1989). 43. B. G. HAFFTY, Y. H. SON. C. T. SASAKI, R. PAPAC, D. FISCHER, S. ROCKWELL, A. SARTORELLI and J. J. FISCHER, Mitomycin C as an adjunct to postoperative radiation therapy in squamous cell carcinoma of the head and neck: results from two randomized clinical trials, Znt. J. Radiar. Oncol. Biol. Phy.s. 27, 241-250 (1993). 44. S. KINOSHITA, K. UZU, K. NAKANO, M. SHIMIZU, T. TAKAHASHI and M. MATSUI, Mitomycin derivatives. I. Preparation of mitosane and mitosene compounds and their biological activities, J. Med. Chem. 14, 103-109 (1971). 45. V. N. IYER and W. SZABALSKI, Mitomycins and porfiromycin: chemical mechanism of activation and cross-linking of DNA. Science 145, 55-58 (1964). 46. S. ROCKWELL, In viva 1 in vitro tumor systems: new models for studying the response of tumors to therapy, Lab. Animal. Sci. 27, 831-851 (1977). 47. S. ROCKWELL, K. A. KENNEDY and A. C. SARTORELLI. Mitomycin C as a prototype bioreductive alkylating agent: in vitro studies of metabolism and cytotoxicity, Int. J. Radiat. Oncol. Biol. Phys. 8, 753-755 (1982).
48. S. ROCKWELL and R. F. KALLMAN, Cellular radiosensitivity and tumor radiation response in the EMT6 tumor cell system, Radiat. Res. 53, 281-294 (1973).
130
A. C. SARTORELLI
et al.
49. S. ROCKWELL, Effects of mitomycin C alone and in combination with X-rays on EMT6 mouse mammary tumors in vivo, J. Natl. Cancer Inst. 71, 765-771 (1983). 50. S. ROCKWELL, Cytotoxicities of mitomycin C and X-rays to aerobic and hypoxic cells in vitro, Int. J. Radiat. Oncol. Biol. Phys. 8, 1035-1039 (1982). 51. S. R. KEYES, S. ROCKWELL and A. C. SARTORELLI, Correlation between drug uptake and selective toxicity of porfiromycin to hypoxic EMT6 cells, Cancer Res. 47, 5654-5657 (1987). 52. N. R. BACHUR, S. L. GORDON and M. V. GEE, Anthracycline antibiotic augmentation of microsomal electron transport and free radical formation, Mol. Phawnacof. 13, 901-910 (1977). 53. R. W. ESTABROOK, Mitochondrial respiratory control and the polarographic measurement of ADP:O ratio, Methods Enzymol. 10, 41-47 (1967). 54. C. A. PRITSOS, P. 0. CONSTANTINIDES, T. R. TRITTON, D. C. HEIMBROOK and A. C. SARTORELLI, Use of high performance liquid chromatography to detect hydroxyl and superoxide radicals generated from mitomycin C, Anal. Biochem. 150, 294-299 ( 1986). 55. S. ROCKWELL, S. R. KEYES, R. LOOMIS, M. KELLEY, D. M. VYAS, H. WONG, T. W. DOYLE and A. C. SARTORELLI, Activity of C-7 substituted cyclic acetal derivatives of mitomycin C and porfiromycin against hypoxic and oxygenated EMT6 carcinoma cells in vitro and in vivo, Cancer Commun. 3, 191-198 (1991). 56. J. E. MOULDER and S. ROCKWELL, Hypoxic fractions of solid tumors: experimental techniques, methods of analysis, and a survey of existing data, lnt. J. Radial. Oncol. Biol. Phys. 10, 695-712 (1984). 57. B. G. HAFFTY, Y. H. SON, M. MOINI, R. PAPAC, D. FISCHER, S. ROCKWELL, A. C. SARTORELLI and J. J. FISCHER, Porfnomycin as an adjunct to radiation therapy in squamous cell carcinoma of the head and neck: results of a Phase I clinical trial, Radial. Oncol. Invest. 1, 297-307 (1994).
Pergamon
Regul., Vol. 35, pp. 117-130, 1995 Copyright 0 1995 Elsevier Science Ltd Printed in Great Britain. All rights reserved Ol%S-2571/951$29.(Xl Enzyme
00652571(94)00015-S
PREFERENTIAL KILL OF HYPOXIC EMT6 MAMMARY TUMOR CELLS THE BIOREDUCTIVE ALKYLATING AGENT PORFIROMYCIN
BY
ALAN C. SARTORELLI, MICHAEL F. BELCOURT, WILLIAM F. HODNICK, SUSAN R. KEYES*, CHRIS A. PRITSOS and SARA ROCKWELL Departments of Pharmacology and Therapeutic Radiology and Developmental Therapeutics Program, Yale Cancer Center, Yale University School of Medicine, New Haven, CT 06520 INTRODUCTION
Solid tumors of both humans and animals contain cells which are distal to the tumor vasculature and consequently exist in an environment of nutritional deficiency, low pH and hypoxia (l-4). Hypoxic cells constitute a therapeutically resistant component of solid tumors. Since oxygen sensitizes cells to the cytotoxic action of X-rays, these cellular elements are resistant to ionizing radiation (1). They are also relatively insensitive to many antineoplastic agents for a variety of reasons (5-12). These include the fact that (a) a number of cancer chemotherapeutic agents have difficulty diffusing into hypoxic regions of tumors in therapeutically adequate concentrations, (b) hypoxic cells progress slowly through the cell cycle or are blocked in their transition through the cycle and, therefore, are resistant to cycle active anticancer drugs, and (c) hypoxia produces amplification of genes capable of conferring resistance to a variety of agents. The oxygen deficiencies within the hypoxic regions of solid tumors lead to an environment conducive to reductive events, thereby creating an opportunity to preferentially attack hypoxic cells with agents that require enzymatic reduction to create electrophilic species. We and others have employed this concept, initially delineated in 1972 (13), to demonstrate that the mitomycin antibiotics preferentially kill hypoxic tumor cells relative to their aerobic counterparts (5-7, 14-18). The cytotoxicity of the mitomycin antibiotics requires bioactivation of these agents to *Present address: Biomeasure, Inc., 27 Maple Street, Mifford, MA 01757-3650. +Present address: Department of Nutrition, University of Nevada, Rem), NV 89557 117
118
A. C. SARTORELLI
et
al.
produce mono- and b&DNA adducts (19-21). A number of enzymes have been demonstrated to be capable of reductively activating the mitomycins, including NADPH:cytochrome P-450 oxidoreductase (EC 1.6.2.4) (22-30), NADH:cytochrome b, oxidoreductase (EC 1.6.2.2) (31, 32), NAD(P)H: quinone oxidoreductase (EC 1.6.99.2; DT-diaphorase) (33-38), xanthine: oxygen oxidoreductase (EC 1.2.3.2; xanthine oxidase) (24, 29, 39), and xanthine: NAD+ oxidoreductase (EC 1.2.1.37; xanthine dehydrogenase) (40, 41). Mitomycin C has been considered by our laboratories to be the prototype bioreductive alkylating agent. We have characterized this agent extensively, demonstrating in mice bearing established solid tumors that this antibiotic in combination with X-irradiation produced additive or greater-than-additive cytotoxicity to tumor tissue, without increased toxicity to normal tissues (for appropriate references see 14). The antineoplastic efficacy of this combination was considered to result from the preferential kill of aerobic tumor cells by the ionizing radiation and of hypoxic cells by mitomycin C. These findings led to the conduct of two prospective randomized clinical trials at Yale University, evaluating mitomycin C as an adjunct to radiation therapy in patients with head and neck cancer; these studies demonstrated significantly improved local control and therapeutic gain with the combination (42, 43). These results have led us to look for agents with greater differential kill of hypoxic and aerobic cells, which might be employed to minimize toxicity to oxygenated host tissues and to maximize the destruction of the hypoxic cells of solid tumors. Porfiromycin has been found to be such an agent, being essentially equivalent to mitomycin C on a molar basis in its toxicity to hypoxic cells in vitro, but being much less toxic than mitomycin C to aerobic cells (15-18,44,45). The present report summarizes data from our laboratories on the effects of porfiromycin on hypoxic and aerobic EMT6 murine mammary tumor cells. MATERIALS
AND
METHODS
Porfiromycin, [sH]porliromycin and mitomycin C were gifts from Dr Terrance Doyle of Bristol Laboratories (Wallingford, CT). NADPH:cytochrome P-450 oxidoreductase was purified from phenobarbital-treated rabbit liver as described previously (28). All other materials were purchased from commercial sources: gases (Presto Welding Service Center, North Haven, CT); fetal bovine serum and Waymouth’s medium (Grand Island Biological Co., Grand Island, NY); and NADH and NADPH (Sigma Chemical Co., St Louis, MO). All experiments were performed using EMT6 mouse mammary tumor cells, which were maintained by alternate passage in vivo and in vitro as
HYPOXIC
CELL ACTIVITY
OF PORFIROMYCIN
119
detailed elsewhere (46). Cell culture studies were performed in Waymouth’s medium supplemented with 15% fetal bovine serum and antibiotics, using exponentially growing monolayers in glass milk dilution bottles. Hypoxia was produced by gassing cultures with a humidified mixture of 95% N,--5% CO, for 2 hr. Mitomycin C and porfiromycin were dissolved in 70% ethanol and small vols of drug were added to the hypoxic cultures without compromising the hypoxia, as described previously (47). The viability of the cells was assayed by measuring the ability of individual cells to form colonies (46). Each experiment included controls subjected to all experimental manipulations and incubated with the vehicle used to dissolve the drug under appropriate aerobic and hypoxic conditions. In vivo studies were performed using EMT6 tumors in 2.5- to 3-monthold BALB/cRw mice by previously published procedures (46,48). Tumors were implanted by the injection of 2 x 105 cells into the skin of the flank and were allowed to grow for 2 weeks, producing tumors approximately 100 cu mM in vol. Porfiromycin was administered i.p. as a single dose 1 hr before sacrifice. Mice were whole body irradiated, without anesthesia, with 15 Gy of 250 kV X-rays at a dose rate of approximately 2 Gy/min (17, 49). Radiation was begun 50 min after the drug treatment. Previous studies demonstrated that the toxicities of mitomycin C and radiation ipr vitro were independent and did not depend upon the sequence or timing of the mitomycin C and radiation treatments (50) and that intervals of 5 min to 24 hr between mitomycin C and X-rays in vivo gave similar cytotoxicities (49). Tumors were excised immediately after irradiation or 1 hr after porfiromycin treatment. Tumor cells were suspended by a combination of mechanical and enzymatic disruption, collected by centrifugation, counted and assayed for viability by determining their ability to form colonies in vitro (46, 48, 49). Uptake and efflux of porhromycin were measured using single-cell suspensions in Waymouth’s medium supplemented with 15’% fetal bovine serum (51). These suspensions were prepared from cells which had been grown in monolayer culture and allowed to recover from trypsinization for 2 hr at 37°C in an atmosphere of 95% air-5% CO,. Five-milliliter aliquots containing 5 x 105 cells/ml were placed in glass scintillation vials and gassed through rubber serum stoppers with 95% N,-5% CO, or 95% air-5% CO, for 2 hr at 37°C. [3H]Porfiromycin was added without compromising the hypoxia. After various time intervals, 0.3 ml aliquots were removed and cell-associated radioactivity was measured by centrifugation through oil (sihcone:mineral oil, 84:16, v/v) into 0.5 N perchloric acid. Radiolabel in the cell pellet was ascertained by scintillation spectrometry in Ultrafluor (National Diagnostics, Somerville, NJ). At the end of the 2 hr uptake period, cells were washed twice with warm medium
120
A. C. SARTORELLI
er al.
equilibrated with 95% air-5% CO, and resuspended in medium to the original concentration of 5 x 10s cells/ml. Cell-associated [sH]porfiromycin was then measured as a function of time after resuspension, using the techniques described above. Throughout the entire procedure, cells were shaken to prevent reattachment. For technical reasons, uptake and efflux studies were performed using suspension cultures (51). The cytotoxicity of porfiromycin was similar for EMT6 cells in monolayer and suspension cultures. Calculation of the intracellular concentration of portiromycin used the following parameters: 5.45 nmol = 1.78 a 0.14 X 106 cpm and cell vol = 2.2 pl. Cell sonicates were prepared from exponentially growing monolayer cultures by detaching cells from flasks using 0.05% trypsin in PBS, washing with cold PBS and, subsequently, resuspending them in the appropriate buffer system. The cells in an ice bath were disrupted by three 5-set bursts with a Branson Model 160 sonicator set at 25% of maximum intensity. Cellular disruption was verified microscopically. Oxygen consumption was used as a measure of oxygen radical generation (52) and was monitored polarographically with a Gilson oxygraph equipped with a Clark electrode (53). The enzymatic activation of drug-induced oxygen consumption was assayed by resuspending EMT6 cells in cold PBS at a concentration of 3-5 x 106 cells/ml and disrupting them by sonication as described above. Final reaction mixtures consisted of 2 ml of EMT6 cell sonicate, 50 mM Tris-HCl buffer (pH 7.8), 1 mM diethylenetriaminepentaacetic acid, 0.5 mM NADPH, 0.5 mM NADH and drug or ethanol as a control. Drugs were dissolved in 70% ethanol to a final concentration of either 1 or 5 mM; dilutions of these stock solutions resulted in no greater than a 1% final concentration of ethanol in incubation flasks, a level that did not affect oxygen consumption. Assays were conducted at 37°C. Initially, cell sonicates and substrates were added to the respiratory cell and oxygen consumption was monitored to obtain a baseline rate for cell sonicate respiration. Once a baseline rate was established, either drug (in ethanol, O-l% final concentration) or ethanol (as a vehicle control) was added and the new rate of oxygen consumption was ascertained. The drug-induced rate of oxygen consumption was obtained by subtracting the rate prior to drug addition from the rate following its addition. This method of determining drug-induced oxygen consumption was used for all of the enzyme assays. Drug-induced oxygen consumption in the presence of NADPH-cytochrome P-450 oxidoreductase was conducted by incubation in 2 ml of 50 mM phosphate buffer (pH 7.4), 0.5 mM NADPH, 0.18 unit of NADPHcytochrome P-450 oxidoreductase and either drug or vehicle; assays were conducted at 30°C.
HYPOXIC
CELL ACTIVITY
OF PORFIROMYCIN
121
The detection of superoxide and hydroxyl radical generated as a consequence of drug activation in sonicates of EMT6 cells was performed using the HPLC procedure described previously (54). This qualitative assay is useful in identifying the oxygen radical species generated by biological systems and can be used to compare relative radical production by various agents in the same system. Prior to sonication, EMT6 cells were suspended at a concentration of 1.6-2.0 X 107 cells/ml in 50 mM Tris-HCI buffer (pi-l 7.8) with 1 mM diethylenetriaminepentaacetic acid. The reaction mixture consisted of 200 ~1 of cell sonicate, 100 mM 5,5-dimethyl-1-pyrroline N-oxide, 0.5 mM NADH, 0.5 mM NADPH and 4.0% ethanol or drug (in ethanol) in a final vol of 250 ~1. The reaction mixture was incubated for 6 min at 37”C, at which time 25 ~1 of the solution were injected into the HPLC system, which consisted of an Altex ultrasphere 3 p,M octadecylsilane 7.5 x 4.6 mm inside diameter column attached to a Bioanalytical Systems LC4 A electrochemical detector (Bioanalytical Systems, Inc., West Lafayette, IN). The detector potential was set at +0.6 V versus an Ag+.-AgCl reference electrode. A Rainin C-18 Microsorb Short-Ones column (Rainin Instrument Co., Woburn, MA) served as a guard column. The mobile phase consisted of 0.03 M citric acid (monohydrate), 0.05 M anhydrous sodium acetate. 0.05 M sodium hydroxide and 8.5% acetonitrile, brought to pH 5.1 with glacial acetic acid; the flow rate was 1.0 ml/min. The relative peak areas of the radicals generated by each drug were determined by cutting the appropriate peaks from the HPLC tracings and weighing them on an analytical balance. RESULTS
AND
DISCUSSION
The toxicities of mitomycin C and porfiromycin (Fig. 1) to hypoxic and aerobic EMT6 cells are shown in Figure 2. On a molar basis, the toxicities of porfiromycin and mitomycin C to EMT6 cells were essentially the same under hypoxia. In contrast, although both mitomycin C and porfiromycin were less toxic to cells under aerobic conditions, porfiromycin was considerably less cytotoxic than mitomycin C in air. Thus, even though the redox potentials of these two mitomycin antibiotics are the same, the addition of a methyl group onto the aziridyl nitrogen (Nla) of mitomycin C markedly decreases its aerobic cytotoxicity, without altering its toxicity under hypoxia. The relatively large difference in the sensitivities of hypoxic and oxgenated EMT6 cells to porfiromycin correlated with differences in the uptake of this antibiotic under these different conditions of oxygenation. Under both aerobic and hypoxic conditions, the rates of uptake of porfiromycin increased with increasing concentrations of radiolabeled drug (Table 1). Uptake consisted of two phases: a rapid phase, which
122
A. C. SARTORELLI
et al.
0 ,4
NH2
**II
3
N
a3 )k!sii
0
Mitomycin C
Porfllycin
FIG. 1. Structures of mitomycin C and porfiromycin.
occurred within 2 min, and a relatively slow phase, which occurred at longer times (51). Uptake occurring during the rapid phase was inhibited by approximately 25-30% when the experiments were performed at 0°C whereas the slower uptake was completely abolished at 0°C. The rapidly associating porfrromycin was blocked approximately 25% by excess unlabeled porfiromycin. The oxygenation state of the cells did not influence the rapid uptake of porfiromycin. These findings suggest that the rapid uptake of porfiromycin primarily represents non-specific binding of the
1
g
10-l
g
.
M .9 .z t
10
-2
I 10-J v) IO4 10-S 0.01
0.1
1 10 Concentration (p&d)
100
FIG. 2. Survival of EMT6 mouse mammary tumor cells exposed to different concentrations of mitomycin C or porlkomycin for 1 hr under hypoxic or aerobic conditions in vitro. Points are geometric means of 240 independent determinations of surviving fractions. The SEMs are shown where n 3 3 and where the error is larger than the point size. Key: Mitomycin C-hypoxic (A), mitomycin C-aerobic (A), porfiromycin-hypoxic (0), porfiromycin-aerobic (0). Reproduced from Ref. (55).
HYPOXIC
CELL ACIWITY
OF PORFIROMYCIN
123
TABLE 1. RATES OF PORFIROMYCIN UPTAKE INTO EMT6 CELLS IN AIR AND IN HYPOXIA
Condition Air Hypoxia
Concentration (PM) 3 10 20 0.2 3
Rate (cpmlhril.5 X 105 cells) 120 300 1100 x0 1000
Reproduced from Ref. (51)
drug. To achieve comparable rates of uptake during the slow phase, a lo-fold lower concentration of porfiromycin was needed in hypoxia than was required in air (Table 1). This difference in drug concentration approximated the difference in porfiromycin levels needed to produce equivalent cytotoxicities under aerobic and hypoxic conditions (51). To determine whether porfiromycin possesses the appropriate physicochemical properties to diffuse through the tumor mass and reach hypoxic cells in concentrations adequate to produce cytotoxicity, the response of solid EMT6 tumors in viva to this antibiotic was measured and the results are shown in Figure 3. The survival curve for tumor cells
Porfkomycin dose (&g) FIG. 3. Effect of varying doses of porfiromycin alone (0) or in combination with 1.5 Gy of X-irradiation (0) on the survival of cells in solid EMT6 tumors. Points: Means + SEMs; 3-8 determinations. Stippled area: Area of additive cytotoxicities calculated from the complete single-agent dose-response curves. Reproduced from Ref. (17).
124
A. C. SARTORELLI
et al.
explanted from mice treated with porfiromycin alone had a very shallow slope. This reflects the fact that approximately 80% of the viable cells in EMT6 tumors are aerobic (56); the overall response of these tumors to porfrromycin is therefore dominated by the response of the aerobic tumor cells. Aerobic tumor cells in vim, like those in vitro (Fig. 2), were relatively resistant to porfiromycin. To examine the effects of porfiromycin on hypoxic tumor cells in vim, we selectively depleted the aerobic tumor cell population by treating the tumors with a large dose of X-rays (15 Gy). Because hypoxic cells are exceedingly resistant to radiation, while aerobic cells are radiosensitive (48, 56), this treatment reduces the survival of the aerated tumor cell population to approximately O.Ol%, while approximately 20% of the hypoxic tumor cells survive (56). Therefore, the tumor cell population surviving irradiation is composed almost entirely of hypoxic cells. The effects of porfiromycin on hypoxic (radioresistant) tumor cells were greater than the effects on the overall tumor cell population. Isobologram analyses, performed using complete dose-response curves for porfiromycin alone and for radiation alone (14, 17,49), indicated that the combinations of 15 Gy of X-rays with 6 ug or 20 pg of porfiromycin per g of body weight produced surviving fractions that lay outside the area of additivity and were compatible with supra-additive effects. This probably reflects the fact that each agent is effective against a tumor cell population which is resistant to the other agent. These findings have led to a Phase I clinical trial of porfiromycin in combination with radiation therapy in patients with advanced head and neck cancer (57). The lower toxicity of porfiromycin to oxygenated cells permitted the use of a porfiromycin dose that was more than 2.5 times the dose used for mitomycin C, with no significant increase in drug toxicity (42, 43, 57). Of 21 patients treated in this series, ten remain alive with no evidence of disease, with a median follow-up of 18.5 months. These results have led to an ongoing Phase III trial at Yale University in patients with squamous cell carcinoma of the head and neck region, comparing long-term local control in patients receiving radiation therapy plus porfiromycin to patients receiving radiation therapy plus mitomycin C. The major difference in the cytotoxicities of porfiromycin to hypoxic and oxygenated EMT6 cells has prompted an evaluation of the importance of the metabolism of the antibiotic in producing this difference in sensitivity. To gain insights into the metabolic processes underlining the hypoxic/aerobic differential, the metabolism of mitomycin C and porfiromycin was compared under aerobic conditions, with the objective being to ascertain the reason for the difference between the aerobic cytotoxicities of these two agents. To ascertain whether mitomycin C and porfiromycin were capable of being activated to a state that generates oxygen radicals in a complete tumor cell system, drug-induced oxygen consumption was measured in EMT6
HYPOXIC
CELL ACTIVITY
OF PORFIROMYCIN
125
cell sonicates supplemented with sufficient NADH and NADPH to ensure that these substrates were not rate-limiting. Both of the compounds tested produced linear concentration-dependent rates of oxygen consumption in the presence of EMT4 cell sonicates and substrate (Fig. 4). The rate of oxygen consumption with mitomycin C was approximately twice that with porfirom ycin . The quantities of superoxide and hydroxyl radical generated by mitomycin C and porfiromycin in sonicates of EMT6 cells were measured and are compared in Table 2. The amounts of superoxide and hydroxyl radicals produced in the presence of mitomycin C were signifkantly greater than those produced by control sonicates. In contrast, radical production in the presence of porfiromycin did not differ from control values. These results presumably are a reflection of the greater capacity of one or more of the enzyme systems present in EMT6 cells to reductively activate mitomycin C, relative to porfiromycin, in air. To ascertain whether NADPH:cytochrome P-450 oxidoreductase was involved in the difference in oxygen consumption between mitomycin C and porfiromycin in sonicates of EMT6 cells, the rate of activated drug-induced oxygen consumption produced by this enzyme was measured and found to be identical for mitomycin C and porfiromycin (Fig. 5). The finding of no difference in the capacity of NADPHcytochrome P-450 oxidoreductase to reductively activate mitomycin C and portiromycin
FIG. 4. Mitomycin antibiotic-induced stimulation of oxygen consumption catalyzed by EMT6 cell sonicates. Reaction mixtures consisted of 2 ml of EMT6 celf aonicates (3-5 x 106 c&s/ml in 50 rnM Tris-HCl buffer, pH 7.8). 1 mM d~e~~~~~~~~c acid, 0.5 mu NADPH and 0.5 mM NADH. Drugs were added at the in#c&?d fintll conooatrations and the reactions were run at 37°C as previously described. 0, mitomycin C; a, portiromycin. Points: Average of at least four determinations: bars, SD. Reproduced from Ref. (25).
126
A. C. SARTORELLI
et al.
TABLE 2. RELATIVE QUANTITIES OF DRUG-INDUCED OXYGEN RADICALS FORMED IN EMT6 CELL SONICATES Compound
Relative production Superoxide Hydroxyl radical
Mitomycin C Porfiromycin
1.29 + 0.03* 0.97 2 0.20
1.18 zk 0.10’ 1.06 + 0.10
*Values with a significant difference of at least P < 0.05 (by the paired Student’s r test) when compared to control samples equated to 1.0. Each value represents the mean + SD from at least 5 determinations. Reproduced from Ref.(25)
makes NADH: cytochrome b, oxidoreductase, DT-diaphorase, xanthine oxidase and/or xanthine dehydrogenase candidate enzymes to explain the difference in the activation of these drugs under aerobic conditions. SUMMARY
Hypoxic cells in solid tumors represent a therapeutically resistant population that limits the curability of many solid tumors by irradiation and by most chemotherapeutic agents. The oxygen deficit, however, creates an environment conducive to reductive processes; this results in a major exploitable difference between normal and neoplastic tissues.
FIG. 5. Mitomycin antibiotic-induced stimulation of oxygen consumption catalyzed by NADPH-cytochrome P-450 oxidoreductase. Reaction mixtures consisted of 2 ml of 50 mM phosphate buffer (pH 7.4), 0.18 unit of NADPH-cytochrome c reductase, 1 mM diethylenetriaminepentaacetic acid and 0.5 mM NADPH. Drugs were added at the indicated final concentrations and the reactions were run at WC as previously described. 0, mitomycin C; 0, portIromycin. Points: average of at least four determinations; bars. SD. Reproduced from Ref. (25).
HYPOXIC CELL ACTIVITY
OF PORFIROMYCIN
127
The mitomycin antibiotics can be reductively activated by a number of oxidoreductases, in a process required for the production of their therapeutic effects. Preferential activation of these drugs under hypoxia and greater toxicity to oxygen-deficient cells than to their oxygenated counterparts are obtained in most instances. The demonstration that mitomycin C and porfiromycin, used to kill the hypoxic fraction, in combination with irradiation, to eradicate the oxygenated portion of the tumor, produced enhanced cytodestructive effects on solid tumors in animals has led to the clinical evaluation of the mitomycins in combination with radiation therapy in patients with head and neck cancer. The findings from these clinical trials have demonstrated the value of directing a concerted therapeutic attack on the hypoxic fraction of solid tumors as an approach toward enhancing the curability of localized neoplasms by irradiation.
ACKNOWLEDGEMENTS
This research was supported in part by Grants DHP-102B (A.C.S.) and EDT-62N (S.R.) from the American Cancer Society.
REFERENCES 1. J. E. MOULDER and S. ROCKWELL, Tumor hypoxia: its impact in cancer therapy, Cancer Metastasis Rev. 5, 313-341 (1987). 2. P. VAUPEL, F. KALLINOWSKI and P. OKUNIEFF, Blood Bow, oxygen, and nutrient suppIy, and metabolic microenvironment of human tumors: a review, Cancer Res. 49, 6449-6465 (1989). .3. J. E. MOULDER, J. DUTREIX, S. ROCKWELL and D. W. SIEMANN, Applicability of animal tumor data to cancer therapy in humans, Inr. J. Radiut. Oncol. Biol. Phys. 14, 913-927 (1988). 4. H. D. SOSTMAN, H. C. CHARLES, S. ROCKWELL, K. LEOPOLD, C. BEAM, D. MADWED, M. DEWHIRST, G. CONFER, D. MOORE, R. BURN and J. OLESON, Soft-tissue sarcomas: detection of metabolic heterogeneity with P-31 MR spectroscopy, Radiology 176, 837-843 (1990). 5. K. A. KENNEDY, B. A. TEICHER, S. ROCKWELL and A. C. SARTORELLI, The hypoxic tumor cell: a target for selective cancer chemotherapy, Biochem. Phnrmacol. 29, l-8 (1980). 6. S. ROCKWELL, Hypoxic cells as targets for cancer chemotherapy, pp. 157-172 in Development of Target-oriented Anticancer Drugs (Y.-C. CHENG, B. GOZ and M. MINKOFF, eds.) Raven Press, New York (1983). 7. A. C. SARTORELLI, Therapeutic attack of hypoxic cells of solid tumors: presidential address, Cancer Res. 48,775-778 (1988). Classification of 8. B. A. TEICHER, J. S. LAZO and A. C. SARTORELLl. antineoplastic agents by their selective toxicities toward oxygenated and hypoxic tumor ceils, Cancer Res. 41, 73-81 (1981). 9. I. F. TANNOCK, The relation between cell proliferation and the vascular system in a transplanted mouse mammary tumour, Brif. J. Cancer 22, 258-273 (I!%@. 10. L. SIMPSON-HERREN and P. E. NOKER, Diversity of penetratton of anticancer agents into solid tumors, Cell Proliferation 24, 355-365 (1991).
128
A. C. SARTORELLI
et al.
11. G. C. RICE, C. HOY and R. T. SCHIMKE, Transient hypoxia enhances the frequency of dihydrofolate reductase gene amplification in Chinese hamster ovary cells, Proc. Natl. Acad. Sci. USA 83, 5978-5982 (1986). 12. G. C. RICE, V. LING and R. T. SCHIMKE, Frequencies of independent and simultaneous selection of Chinese hamster cells for methotrexate and doxorubicin (adriamycin) resistance, Proc. Natl. Acad. Sci. USA 84, 9261-9264 (1987). 13. A. J. LIN, L. A. COSBY, C. W. SHANSKY and A. C. SARTORELLI, Potential bioreductive alkylating agents. I. Benzoquinone derivatives, J. Med. Chem. 15, 1247-1252 (1972). 14. S. ROCKWELL and A. C. SARTORELLI, Mitomycin C and radiation, pp. 125-139 in Interactions Between Antitumor Drugs and Radiation (B. T. HILL and A. S. BELLAMY, eds.) CRC Press, New York (1990). 15. S. R. KEYES, S. ROCKWELL and A. C. SARTORELLI, Porfnomycin as a bioreductive alkylating agent with selective toxicity to hypoxic EMT6 tumor cells in vivo and in vitro, Cancer Res. 45, 3642-3645 (1985). 16. P. M. FRACASSO and A. C. SARTORELLI, Cytotoxicity and DNA lesions produced by mitomycin C and porfiromycin in hypoxic and aerobic EMT6 and Chinese hamster ovary cells, Cancer Res. 46, 3939-3944 (1986). 17. S. ROCKWELL, S. R. KEYES and A. C. SARTORELLI, Pre-clinical studies of porfiromycin as an adjunct to radiotherapy, Radiat. Res. 116, 100-113 (1988). 18. R. S. MARSHALL and A. M. RAUTH, Oxygen and exposure kinetics as factors influencing the cytotoxicity of porfiromycin, a mitomycin C analogue, in Chinese hamster ovary cells, Cancer Res. 48, 5655-5659 (1988). 19. S.-S. PAN, Porfiromycin disposition in oxygen-modulated P388cells, Cancer Chemother. Pharmacol. 27, 187-193 (1990). 20. M. TOMASZ, C. S. HUGHES, D. CHOWDARY, S. R. KEYES, R. LIPMAN, A. C. SARTORELLI and S. ROCKWELL, Isolation, identification, and assay of [3H]-porfiromycin adducts of EMT6 mouse mammary tumor cell DNA: effects of hypoxia and dicumarol on adduct patterns, Cancer Commun. 3, 213-223 (1991). 21. R. BIZANEK, D. CHOWDARY, H. ARAI, M. KASAI, C. S. HUGHES, A. C. SARTORELLI, S. ROCKWELL and M. TOMASZ, Adducts of mitomycin C and DNA in EMT6 mouse mammary tumor cells: effects of hypoxia and dicumarol on adduct patterns, Cancer Res. 53. 5127-5134 (1993). 22. K. KENNEDY, S. G. SLIGAR; L. POLOMSKI and A. C. SARTORELLI, Metabolic activation of mitomycin C by liver microsomes and nuclei, Biochem. Pharmacol. 11, 2011-2016 (1982). 23. T. KOMIYAMA, T. OK1 and T. INUI, Activation of mitomycin C and quinone drug metabolism by NADPH-cytochrome P-450 reductase, J. Pharm. Dyn. 2, 407410 (1979). 24. N. R. BACHUR, S. L. GORDON, M. V. GEE and H. KON, NADPH cytochrome P-450 reductase activation of quinone anticancer agents to free radicals, Proc. Natl. Acad. Sci. USA 76, 954-957 (1979). 25. C. A. PRITSOS and A. C. SARTORELLI, Generation of reactive oxygen radicals through bioactivation of mitomycin antibiotics, Cancer Res. 46, 3528-3532 (1986). 26. T. KOMIYAMA, T. KIKUCHI and Y. SUGIURA, Generation of hydroxyl radical by anticancer quinone drugs, carbazilquinone, mitomycin C, aclacinomycin A and adriamycin, in the presence of NADPH-cytochrome P-450 reductase, Biochem. Pharmacol. 31, 3651-3656 (1982). 27. S.-S. PAN, P. A. ANDREWS and C. J. GLOVER, Reductive activation of mitomycin C metabolites catalyzed by NADPH-cytochrome P-450 reductase and xanthine oxidase, J. Biol. Chem. 259, 959-966 (1984). 28. S. R. KEYES, P. M. FRACASSO, D. C. HEIMBROOK, S. ROCKWELL, S. G. SLIGAR and A. C. SARTORELLI, Role of NADPH:cytochrome C reductase and DT-diaphorase in the biotransformation of mitomycin C, Cancer Res. 44, 5638-5643 (1984). 29. S.-S. PAN, T. IRACKI and N. R. BACHUR, DNA alkylation by enzyme-activated mitomycin C, Mol. Pharmacol. 29, 622-628 (1986).
HYPOXIC CELL ACTIVITY
OF PORFIROMYCIN
129
30. H. F. J. BLIGH, A. BARTOSZEK, C. N. ROBSON, I. D. HICKSON, C. B. KASPER, J. D. BEGGS and C. R. WOLF, Activation of mitomvcin C bv NADPH: cytocbrome P-450 reductase, Cancer Res. 50, 7789-7792 (1990). _ . 31. W. F. HODNICK and A. C. SARTORELLI, Reductive activation of mitomycin C by NADH:cytochrome bs reductase, Cancer Res. 53,4907-4912 (1993). 32. J. FISHER and R. OLSEN, Mechanistic aspects of mitomycin C activation by Aavoprotein transhydrogenases, Develop. Biochem. 21, 240-243 (1982). 33. C. A. PRITSOS, L. L. PARDINI, A. J. ELLIOTT and R. S. PARDINI, Relationship between the antioxidant enzyme DT-diaphorase and tumor response to mitomycin C treatment, pp. 713-716 in Oxygen Radicals in Biology and Medicine (M. G. SIMIC, K. A. TAYLOR, J. F. WARD and C. VON SONNTAG. eds.) Plenum. New York (1989). 34. A. BEGLEITER, E. ROBOTHAM, G. LACEY and M. K. LEITH, Increased sensitivity of quinone resistant cells to mitomycin C. Cancer L,effers 49, 373-176 (1989). 35. A. BEGLEITER, E. ROBOTHAM and M. K LEITH, Role of NAD(P)H:(quinone acceptor) oxidoreductase (DT-diaphorase) in activation of mitomycin C under hypoxia, Mol. Pharmacol.
41, 677-682 (1992).
36. D. SIEGEL, N. W. GIBSON, P. C. PREUSCH and D. ROSS, Metabolism of mitomycin C by DT-diaphorase: role in mitomycin C-induced DNA damage and cytotoxicity in human colon carcinoma cells, Cancer Res. SO, 7483-7489 (1990). 37. D. SIEGEL, H. BEALL, C. SENEKOWITSCH, M. KASAI, H. ARAI, N. W. GIBSON and D. ROSS, Bioreductive activation of mitomycin C by DT-diaphorase, Biochemistry. 31, 7879-7885 (1992). 38. H. D. BEALL, R. T. MULCAHY, D. SIEGAL, R. D. TRAVER, N. W. GIBSON and D. ROSS, Metabolism of bioreductive antitumor compounds by purified rat and human DT-diaphorases, Cancer Res. 54, 3196-3201 (1994): 39. K. A. KENNEDY, E. G. MIMNAUGH, M. A. TRUSH and B. K. SINHA, Effects of glutathione and ethylxanthate on mitomycin C activation by isolated rat hepatic or EMT6 mouse mammary tumor nuclei, Cancer Res. 45, 4071-4076 (1985). 40. D. L. GUSTAFSON and C. A. PRITSOS, Bioactivation of mitomycin C by xanthine dehydrogenase from EMT6 mouse mammary carcinoma tumors, J. Nat/. Cuncer Insf. 84, 1180-1185 (1992). 41. D. L. GUSTAFSON and C. A. PRITSOS, Enhancement of xanthine dehydrogenase mediated mitomycin C metabolism by dicumarol, Cancer Res. 52, 6936-6939 (1992). 42. J. B. WEISSBERG, Y. H. SON, R. J. PAPAC, C. SASAKI, D. B. FISCHER, R. LAWRENCE, S. ROCKWELL, A. C. SARTORELLl and J. J. FISCHER, Randomized clinical trial of mitomycin C as an adjunct to radiotherapy in head and neck cancer. Int. J. Radial. Oncol. Biol. Phys. 17, 3-9 (1989). 43. B. G. HAFFTY, Y. H. SON. C. T. SASAKI, R. PAPAC, D. FISCHER, S. ROCKWELL, A. SARTORELLI and J. J. FISCHER, Mitomycin C as an adjunct to postoperative radiation therapy in squamous cell carcinoma of the head and neck: results from two randomized clinical trials, Znt. J. Radiar. Oncol. Biol. Phy.s. 27, 241-250 (1993). 44. S. KINOSHITA, K. UZU, K. NAKANO, M. SHIMIZU, T. TAKAHASHI and M. MATSUI, Mitomycin derivatives. I. Preparation of mitosane and mitosene compounds and their biological activities, J. Med. Chem. 14, 103-109 (1971). 45. V. N. IYER and W. SZABALSKI, Mitomycins and porfiromycin: chemical mechanism of activation and cross-linking of DNA. Science 145, 55-58 (1964). 46. S. ROCKWELL, In viva 1 in vitro tumor systems: new models for studying the response of tumors to therapy, Lab. Animal. Sci. 27, 831-851 (1977). 47. S. ROCKWELL, K. A. KENNEDY and A. C. SARTORELLI. Mitomycin C as a prototype bioreductive alkylating agent: in vitro studies of metabolism and cytotoxicity, Int. J. Radiat. Oncol. Biol. Phys. 8, 753-755 (1982).
48. S. ROCKWELL and R. F. KALLMAN, Cellular radiosensitivity and tumor radiation response in the EMT6 tumor cell system, Radiat. Res. 53, 281-294 (1973).
130
A. C. SARTORELLI
et al.
49. S. ROCKWELL, Effects of mitomycin C alone and in combination with X-rays on EMT6 mouse mammary tumors in vivo, J. Natl. Cancer Inst. 71, 765-771 (1983). 50. S. ROCKWELL, Cytotoxicities of mitomycin C and X-rays to aerobic and hypoxic cells in vitro, Int. J. Radiat. Oncol. Biol. Phys. 8, 1035-1039 (1982). 51. S. R. KEYES, S. ROCKWELL and A. C. SARTORELLI, Correlation between drug uptake and selective toxicity of porfiromycin to hypoxic EMT6 cells, Cancer Res. 47, 5654-5657 (1987). 52. N. R. BACHUR, S. L. GORDON and M. V. GEE, Anthracycline antibiotic augmentation of microsomal electron transport and free radical formation, Mol. Phawnacof. 13, 901-910 (1977). 53. R. W. ESTABROOK, Mitochondrial respiratory control and the polarographic measurement of ADP:O ratio, Methods Enzymol. 10, 41-47 (1967). 54. C. A. PRITSOS, P. 0. CONSTANTINIDES, T. R. TRITTON, D. C. HEIMBROOK and A. C. SARTORELLI, Use of high performance liquid chromatography to detect hydroxyl and superoxide radicals generated from mitomycin C, Anal. Biochem. 150, 294-299 ( 1986). 55. S. ROCKWELL, S. R. KEYES, R. LOOMIS, M. KELLEY, D. M. VYAS, H. WONG, T. W. DOYLE and A. C. SARTORELLI, Activity of C-7 substituted cyclic acetal derivatives of mitomycin C and porfiromycin against hypoxic and oxygenated EMT6 carcinoma cells in vitro and in vivo, Cancer Commun. 3, 191-198 (1991). 56. J. E. MOULDER and S. ROCKWELL, Hypoxic fractions of solid tumors: experimental techniques, methods of analysis, and a survey of existing data, lnt. J. Radial. Oncol. Biol. Phys. 10, 695-712 (1984). 57. B. G. HAFFTY, Y. H. SON, M. MOINI, R. PAPAC, D. FISCHER, S. ROCKWELL, A. C. SARTORELLI and J. J. FISCHER, Porfnomycin as an adjunct to radiation therapy in squamous cell carcinoma of the head and neck: results of a Phase I clinical trial, Radial. Oncol. Invest. 1, 297-307 (1994).