[CANCER RESEARCH 55, 574-580, February 1, 1995]

/Y.[2-(2-Methyl-5-nitroimidazolyl)ethyl]-4-(2-nitroimidazolyl)butanamide

(NSC 639862), a Bisnitroimidazole with Enhanced Selectivityas a Bioreductive Drug1

John W. Moselen, Michael P. Hay, William A. Denny, and William R. Wilson2

Section of Oncology, Department of Pathology ¡J.W. M., W. R. W.J and Cancer Research Laboratory [M. P. H., W. A. DJ, University of Auckland School of Medicine, PrivateBag 92019, Auckland, New Zealand

ABSTRACT

Compounds containing two redox centers, both of which must bereduced for full expression of cytotoxicity by oxygen-inhibitable pathways(bis-bioreductive drugs), have potential as cytotoxins with high selectivityfor hypoxic tumor cells. The bisnitroimidazole JV-[2-(2-methyl-5-nitroimi-dazolyl)ethyl]-4-(2-nitroimidazolyl)butanamide (NNB, NSC 639862), inwhich a 2-nitroimidazole and 5-nitroimidazole moiety are joined via a

carboxamide linker, is highly selective for hypoxic AA8 Chinese hamstercells (200-fold by 8 h) relative to mononitroimidazoles (5-25-fold). Abis-bioreductive mechanism is consistent with the marked increase in

hypoxic potency and selectivity of NNB with time and the apparentrequirement that the two nitro groups be present in the same molecule.NNB differed from mononitroimidazoles in inducing fewer DNA singlestrand breaks at equivalent toxicity, suggesting that a duplex DNA lesion(locally doubly damaged site) may be responsible for cell killing. Alkalineelution studies and the lack of hypersensitivity of the repair-defective UV4

cell line indicate that the cytotoxic lesion is not a DNA interstrand crosslink. NNB shows greater hypoxic selectivity than the alkylating 2-nitroi

midazole RB 6145 against AA8 cells and is active in combination withradiation when administered in multiple doses against the MDAH-MCa-4

mouse mammary carcinoma.

INTRODUCTION

There is evidence that many human tumors contain hypoxic cells(1) and that the radioresistance of these cells can be a limiting factorin radiotherapy (2-6). Studies with murine tumors have shown that

nitroimidazoles can partially substitute for oxygen to radiosensitizehypoxic cells (7, 8); some trials in humans have also shown radio-

sensitization by nitroimidazoles (5) but results have generally beendisappointing at clinically achievable doses.

In addition to their radiosensitizing activity, nitroimidazoles areselectively toxic to hypoxic cells (9-11) as a consequence of theirreductive activation via oxygen-inhibitable metabolic pathways. Such"bioreductive drugs" have the potential to improve tumor response to

radiotherapy by killing hypoxic cells (12-14) and might be preferable

to hypoxic cell radiosensitizers inasmuch as they offer potential forturning hypoxia to advantage (15-17). The selective metabolic activation of 2-nitroimidazoles (and resulting covalent binding) in hypoxic regions of tumors is well documented (18-21), but the lowcytotoxic potency of the reduction products of simple 2-nitroimida

zoles precludes their use as bioreductive drugs (22).The alkylating 2-nitroimidazole RSU 1069, however, has an aziri-

dine side chain which confers greater cytotoxic potency and higherselectivity for hypoxic cells (23, 24). Both RSU 1069 and its bromo-

ethylamino prodrug RB 6145 (Fig. 1) are effective in killing hypoxic

Received 7/21/94; accepted 11/29/94.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

1This work was supported by Contract NO1-CM-07321 from the National Cancer

Institute. J. W. M. was the recipient of a scholarship from the Health Research Council ofNew Zealand.

2 To whom requests for reprints should be addressed.

cells in murine tumors (25-27), and the less emetic R enantiomer of

RB 6145 (PD 144872) is currently scheduled for clinical trial.In RB 6145 both the bromoethylamino and nitro group give rise to

DNA-reactive centers, but only the nitro group is activated selectivelyunder hypoxic conditions (28-30). The aziridine moiety, while en

hancing hypoxic cytotoxic potency, is also likely to contribute togastrointestinal toxicity (31). We have therefore investigated replacement of the bromoethylamino alkylating center of RB 6145 with asecond nitroimidazole moiety which will give rise to a reactive centerpreferentially under hypoxic conditions. We have earlier suggestedthat such incorporation of two redox centers into the same molecule("bis-bioreductive drugs") might provide a general approach for en

hancing hypoxic selectivity (14, 32). This approach requires that bothredox centers are reduced by O2-inhibitable pathways and that thebis-reduced species is much more cytotoxic than the monoreduced

intermediates. The latter condition might be met by a bisnitroimidazole if reactive nitroreduction intermediates derived from both nitrogroups can react with DNA to produce a duplex lesion such as aninterstrand cross-link or dsb.3

A series of carboxamide-linked bisnitroimidazoles has recentlybeen synthesized to test this approach (33). Investigation of structure-

activity relationships was constrained by the limited aqueous solubility of these compounds, but two compounds with a 2-nitroimidazolelinked to a 5-nitroimidazole were identified as having a hypoxic

selectivity greater than related mononitroimidazoles (33). In the present study the more soluble of these two compounds, NNB (NSC639862; Fig. 1), is examined to test the hypothesis that it acts as abis-bioreductive agent under hypoxic conditions and to evaluate the

therapeutic potential of this approach. The hypoxic toxicity and mechanism of action of NNB are compared with those of mononitroimidazoles, including the close analogue INB which lacks the 5-nitro

group (Fig. 1).

MATERIALS AND METHODS

Chemicals. MISO was a gift from the National Cancer Institute (Bethesda,MD), METRO was from May & Baker NZ, Ltd. (Lower Hutt, Wellington,New Zealand), and RB 6145 was from Parke-Davis Pharmaceutical Research

Division, Warner Lambert Co. (Ann Arbor, MI). NNB and INB were synthesized as described (33). The purities of these compounds were >99% by HPLCbased on peak areas at 320 nm. All were dissolved in CM and filter sterilizedbefore use. Concentrations were determined by spectrophotometry, usingextinction coefficients in 0.01 N HC1 of 7,750 M'1 cm~' at 326 nm for MISO,5,320 M"1 cm'1 at 280 nm for METRO, 12,480 M'1 cm'1 at 322 nm for NNB,and 4,760 M~' cm"1 at 326 nm for INB. Initial concentrations of RB 6145 were

estimated similarly with reference to a standard solution prepared in CM and

3 The abbreviations used are: dsb, double strand break; CM, culture medium (aMEM

containing 5% FBS plus 100 ID/ml penicillin and 100 /ig/ml streptomycin); C,0 or CTmconcentration or concentration X time to reduce survival to 10% of controls; INB,/V-[2-(imidazolyI)ethyl]-4-(2-nitroimidazolyl)butanamide; LDDS, locally doubly damaged site; METRO, metronidazole; MISO, misonidazole; NNB, /V-[2-(2-methyl-5-nitroi-midazolyl)ethyl]-4-(2-nitroimidazolyI)butanamide; ssb, single strand break; 7"10,time toreduce survival to 10% of control; ['"CJdThd, [5-mcr/iv;-'4C]thymidine; [3H]dThd,[5-mti/n7-1H]thymidine.

574

Research. on February 13, 2020. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

BIS-NITROIMIDAZOLE-BIOREDUCTIVE DRUGS

_^N0-^-M

•OCH3o^~-"M

MISO METRO

OH

RB6U5

C N N

V- J>"W / ^--M\

NNB INB

Fig. 1. Structures of nitroimiiiazoles investigated.

allowed to react to completion (34) (final extinction coefficient. 7,620 M~'

cm"' at 326 nm). [l4C]dThd and pHJdThd were purchased from NEN Re

search Products (Boston. MA) and proteinase K was from Sigma Chemical Co.

(St. Louis, MO).Cell Culture. AA8 (35) and UV4 (36) Chinese hamster cell lines were

passaged as monolayer cultures using CM without antibiotics and grown forexperiments in 250-ml glass spinner flasks in CM with the fetal bovine serumraised to 10%. Log-phase cultures were adjusted to 5 X 10s (AA8) or 3 X IO5(UV4) cells/ml to give, 20 h later, early plateau phase cultures at about I X IO6(AA8) and 5 X IO5 (UV4) cells/ml.

Assessment of Hypoxia-selective Cytotoxicity. Aerobic (20% O2) and

hypoxic cytotoxicity was assessed by clonogenic assay of stirred, continuouslygassed cell suspensions (IO6 cells/ml) as detailed elsewhere (37), with drug

solutions in CM gassed with 5% CO2 in N2 or air for 60 min prior to additionof a concentrated suspension of hypoxic cells. O2 concentrations in solutionwere determined using a Clark-type O2 electrode with a high-stability voltage

supply and amplifier (38). Rates of killing were assessed by fitting the survivaldata with second order regressions to interpolate Tln values. The reproducibil-

ity of the cytotoxicity assay was evaluated by treating hypoxic AA8 cells withNNB at 3.2 mM, providing an intraexperiment mean Ttil of 1.6 h (coefficientof variation 7%) for 4 replicate cultures and an interexperiment mean of 1.7 h(coefficient of variation 6%) in 3 independent experiments.

DNA Elution. Cells for alkaline elution were grown to 5 x IO5(AA8) or3 x IO5 (UV4) cells/ml and labeled with either [l4C]dThd (0.37 kBq/ml, 2.1GBq/mmol) or [3H]dThd (3.7 kBq/ml, 710 MBq/mmol) for 20 h beforecentrifugation to remove unincorporated label. |'4C]dThd-labeled cells were

exposed to drugs for 5 h under hypoxic conditions as above. A sample was thenremoved for determination of cell survival and the remainder were centrifugedand resuspended in ice-cold PBS at 5 X IO5 cells/ml. In some experiments

cells were then -y-irradiated (3 Gy) as stirred aerobic suspensions on ice, usingan Eldorado G 60Co teletherapy unit (Nordion, Ltd., Ottawa, Ontario, Canada)

at a dose rate of 2.2 Gy/min, for determination of DNA cross-links. Non-drug-treated [3H]dThd-labeled cells were irradiated (3 Gy) to provide an internalstandard. Samples containing 5 X IO5 each of 14C- and 'H-labeled cells were

lysed with proteinase K and sodium dodecyl sulfate at pH 10 for 1 h onpolycarbonate filters and eluted at pH 12.1 as described by Kohn et al. (39). ssbfrequencies, expressed relative to radiation, were quantitated by interpolatingthe relative retention, K (fraction of 14C retained on the filter when 50% of theinternal 3H standard had eluted). A standard curve for radiation was determined from two experiments with I4C-labeled AA8 cells irradiated as oxy

genated suspensions on ice. The relationship between log,,,/? and the radiationdose D (in Gy) was linear (log,,,« = -0.11 D; r = 0.99) over the rangeinvestigated (0-7.5 Gy). Nondenaturing elution (pH 9.6) was performed asdescribed by Bradley and Kohn (40); [l4C]dThd-labeled cells were exposed todrug and loaded onto polycarbonate filters as above, but using [•'HjdThd-

labeled cells irradiated to a dose of 75 Gy as the internal standard.Activity against MDAH-MCa-4 Tumors. The MDAH-MCa-4 mouse

mammary carcinoma (41) was obtained from Dr. Helen B. Stone, RadiationOncology Research Laboratory, UCSF, CA, at the second transplant genera

tion. Cell suspensions were prepared from fourth generation s.c. tumors bymincing with scissors, stirring magnetically in CM for 30 min at roomtemperature, filtering through a 200 mesh screen, and collecting the cells in thefiltrate by centrifugation. The pellet was resuspended in CM without serum andmice were given i.m. injections in the right gastrocnemius muscle (5 (j.1packedcell volume) to provide fifth passage isotransplants in female C3H/HeN micefor experiments. The tumor was not significantly immunogenic in this strain,with the packed cell volume required for 50% tumor takes being 0.08 fj.1(95%confidence limits. 0.055-0.11 /xl by logistic regression) in naive mice and 0.23HÌ(95% confidence limits, 0.08-0.62 /xl) in mice immunized by i.p. injection

of lethally irradiated cells (IO (il packed cell volume) weekly for 3 weeksbefore challenging with viable cells in both hind legs (7 mice/group).

For tumor growth delay experiments tumor size was measured by determining the smallest diameter circular hole through which the leg would pass.Mice were allocated to treatment groups (7 mice/group) when the tumor plusleg diameter reached 10 mm (0.5-g tumors). Tumors were irradiated (20 Gy;

Philips RT 250; 175 kVp; filtration, 0.2 mm Cu; 1.05 Gy/min) by restrainingunanesthetized mice in a Lucite-lined box. with the tumor-bearing leg drawn

out and tethered in the radiation field using a wire clip around the ankle. Thebody was shielded by lining the lid of the restraining box with lead; the doseto the center of the body was 3% of that to the tumor as assessed bythermoluminescent chip dosimetry. Drugs were formulated in PBS and administered i.p. at 0.05 ml/g. Differences in the time required to reach a leg diameterof 13 mm were tested by ANOVA followed by Dunnett's test to assess

significance of the growth delay in individual groups.

RESULTS

Cytotoxicity of NNB. The time and concentration dependence ofcell killing by the bisnitroimidazole NNB was investigated usingstirred, continuously gassed cultures of AA8 cells. Its cytotoxicity wasmarkedly enhanced under hypoxic conditions, particularly when cellswere exposed to low drug concentrations for long times (Fig. 2). CTWvalues, calculated as an inverse measure of cytotoxic potency, wereconstant with time under aerobic conditions but decreased progressively under hypoxia (Fig. 3). Hypoxic selectivity (determined as theratio of drug concentrations required for 10% survival under air or N2)thus increased markedly with time to give a ratio of about 200 by 8 h(Fig. 4). The potency and hypoxic selectivity of NNB against theAA8-derived repair-defective UV4 cell line, which is hypersensitiveto agents causing DNA cross-links and bulky monoadducts (42, 43),

was similar to that for AA8 cells (Figs. 3 and 4).Cytotoxicity of RB 6145. A similar investigation of the hypoxic

selectivity of RB 6145 was performed using the AA8 and UV4 cell

Fig. 2. Plating efficiency of AA8 cells exposed to NNB in stirred suspension culturesunder aerobic (O, O, A, V) or hypoxia (•,•A, T. ^) conditions. O, •,control; D, 10mu; A, 30 HIM;V, 40 mM; •.0.05 mm; A, 0.1 min; Y, l mM; *, IO mM.

575

Research. on February 13, 2020. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

BIS-NITROIMIDAZOLE-BIOREDUCTIVE DRUGS

lines. CT10 values were almost constant with time under aerobicconditions but decreased progressively under hypoxia in both celllines (Fig. 3), although this change was less pronounced than withNNB. Hypoxie selectivity increased about 2-fold between 1 and 8 h inthe AA8 cell line and 3-fold in the UV4 line (Fig. 4). In contrast to

NNB, UV4 cells were much more sensitive to RB 6145 than wereAA8 cells [differential 8-fold under aerobic conditions and 20-fold

under hypoxia (Fig. 3)]; hypoxic selectivity was also much greater inthe repair-deficient line (Fig. 4).

Cytotoxicity of Mononitroimidazoles. The mononitroimidazolesMISO, METRO, and INB were investigated using the AA8 cell line(Fig. 3). For MISO and INB the aerobic CTW was again constant withtime; aerobic cell killing by the least potent compound (METRO)could be detected only using long exposure times and a concentration(72 HIM)close to the solubility limit. Under hypoxia there was a slighttrend towards lower CTW values at late times for MISO and INB andthe hypoxic selectivity of these compounds thus increased with time,although much less dramatically than for NNB (Fig. 4). At late times(about 8 h) hypoxic selectivities were 12-fold for INB, 25-fold forMISO, and 5-fold for METRO. All three mononitro compounds were

thus much less potent and selective than NNB as hypoxic cell toxinsat this time.

Variation of Pregassing Time. The large increase in potency ofNNB under hypoxia with time (and the smaller increases observedwith RB 6145, MISO, and INB) raised concerns that residual oxygenmight have been present in the culture system at early times. Measurement of O2 in solution indicated concentrations of about 0.4 \JMin CM (equivalent to 0.04% O2 in the gas phase) after the 60-min

pregassing period. On addition of cells, O2 fell below the limit ofdetection (0.04 /AM)within 10 min, suggesting that complete anoxia is

•00 o—ee-X,<&

o oe

—GB—\—

•—•-••o•

••••—-&-e®-—

•-•—m**10

1 10 1 10 1 10 1 10

Time (h)

Fig. 3. C7",o values versus drug exposure time in aerobic (O, D) and hypoxic (•,•)

cultures of AA8 (O, 0) and UV4 (G, •)cells. Values were determined at a range of drugconcentrations as illustrated for NNB in Fig. 2.

200

•^ 150

100

OO.

50

AA8NNB

INB

200

1 2345678

Time (h)

1 2345678

Time (h)

Fig. 4. Changes in hypoxic selectivity (aerobic Cr10/hypoxic CT10) of nitroimidazoleswith time in AA8 (left) and UV4 (right) cultures. Ratios calculated from the data ofFig. 3.

2.0

1.5

1.0

0.5

0.0

Time before mixing (h)

Fig. 5. Time required for NNB (3.2 mM) to reduce survival of hypoxic AA8 cells to10% (ordinate) as a function of the time for which the drug solution in medium wasequilibrated with 5% CO2 in N2 before addition of cells (abscissa).

achieved quickly. However, the reduction of 5-nitroimidazoles ap

pears to be exquisitely sensitive to traces of O2 (44); therefore thesemeasurements did not definitively exclude the possibility that 02concentrations below the sensitivity limit of the electrode mightinhibit activation of the bisnitroimidazole. Rates of killing by NNB(3.2 mM) were therefore measured after gassing the drug solution forvarying periods before addition of cells. This demonstrated a smallprogressive increase in the rate of killing when the duration ofpregassing was extended beyond the standard time of 1 h, but the Twdecreased by only 30% between 1 and 6 h (Fig. 5). In contrast, thehypoxic CT10 value in the experiments of Fig. 3 decreased by >100-

fold over an equivalent period. Similar experiments with METRO (72mM) indicated a decrease of only 25% in the hypoxic Tw when thegassing time before mixing was increased from 1 to 2 h. Thus thelarge increase in potency of NNB when exposure time is extendeddoes not appear to result from decreasing O2 concentrations.

Cytotoxicity of Equimolar MISO and METRO. The effect ofthe 5-nitroimidazole METRO on the Cytotoxicity of the 2-nitroimida-

zole MISO was determined to assess whether the high hypoxic potency and selectivity of NNB requires that both the nitroimidazolemoieties be present in the same molecule. Equimolar METRO enhanced the cytotoxic potency of MISO (40 HIM)under aerobic conditions but had no effect on the toxicity of MISO (3 mM) underhypoxia (Fig. 6). Thus, an equimolar mixture of these 2- and 5-ni

troimidazoles has even less differential for hypoxic cells than doesMISO alone.

DNA Elution Studies. The DNA ssb frequencies induced by nitroimidazoles in hypoxic AA8 cells were determined by alkalineelution using 5-h hypoxic drug treatments and were compared with

cell killing measured in the same experiments (Fig. 7). The elutioncurves for all compounds were approximately exponential (data notshown), except for RB 6145 which showed a distinct lag, presumablyreflecting slow breakage at alkali-labile sites as reported for RSU

1069 (39). All four mononitroimidazoles induced ssb at broadlysimilar frequencies (equivalent to about 5 Gy radiation at drug concentrations giving 99% cell kill). In contrast, NNB failed to induce ssbsignificantly above that in hypoxic controls (which consistentlyshowed slightly faster DNA elution than cells not incubated underhypoxia), even at concentrations up to 3-fold greater than the highest

in Fig. 7.Treatment of AA8 cells with NNB (up to 2.5 HIM)under hypoxia

for 5 h failed to prevent the increase in DNA elution rate caused by

576

Research. on February 13, 2020. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

BIS-NITROIMIIJAZOI I -HIORhDUCTIVE DRUGS

10

O 1 34567

Time (h)

Fig. 6. Cytotoxicily of equimolar MISO and METRO in aerobic (O, Ü,A, V) andhypoxic (•,•.A, T) AA8 cultures. O, •,control; D, 40 ml«METRO; A, 40 mM MISO;V, 40 mM METRO and 40 mM MISO; •3 mm METRO; A. 3 mM MISO; T, 3 mMMETRO and 3 mM MISO.

O

1 0.1 0.01 0.001 0.0001

Surviving fraction

Fig. 7. DMA single strand break frequency expressed as ionizing radiation equivalentsversus surviving fraction after treatment of AA8 cells with nitroimidazoles under hypoxiafor 5 h. 22, range of single strand break frequencies in AA8 cells incubated under hypoxiafor 5 h without drugs. O, MISO; D, METRO: V, INB; A, RB 6145; O. NNB; , linearregression through NNB data; -—. linear regression through all other compounds.

irradiation (Fig. SA). Thus DNA interstrand cross-links could not be

detected even at these highly cytotoxic drug concentrations. In parallelexperiments using chlorambucil (5 h hypoxic exposure) as a positivecontrol, extensive cross-linking was demonstrated (Fig. 8ß)in thecytotoxic concentration range (D1() = 32 /U.Munder these conditions;

data not shown). Thus at equivalent toxicity NNB was much lesseffective than chlorambucil in inducing DNA cross-links. Similarstudies with the cross-link repair-defective UV4 line using 1 mM NNBunder hypoxia for 5 h also failed to demonstrate cross-links (data not

shown).Nondenaturing elution experiments showed a DNA double strand

break frequency after treatment of hypoxic AA8 cells with NNB (20mM, 5 h) which was equivalent to irradiation of oxygenated cells toabout 30 Gy (data not shown). These doses are 4 and 130 times thoserequired for 90% cell kill with radiation and NNB, respectively,suggesting that the cytotoxic lesions induced by NNB are not radiation-like double strand breaks. However, the relevance of these DNA

breakage data at such toxic drug concentrations and long exposuretimes are unclear.

Activity against Hypoxic Cells in Tumors. The activity of NNBagainst MDAH-MCa-4 mammary tumors was assessed in combina

tion with radiation and compared with MISO. NNB alone had no

significant antitumor activity at a dose of 2.5 mmol/kg and, unlikeMISO, did not significantly increase growth delay when administered30 min before radiation (Fig. 9). This was the highest single dose ofNNB which could be administered because of solubility limitations,but the total dose could be increased without morbidity using amultidose schedule (every 3 h for 4 treatments). This schedule did notimprove the activity of MISO (Fig. 9/4), but irradiation 30 min afterthe last NNB dose provided a statistically significant (P < 0.01)growth delay [11.2 ±3.8 (SE) days] additional to radiation alone (Fig.9ß).The activity of multiple doses of NNB (2.5 mmol/kg, every 3 hfor 4 treatments) in combination with radiation was confirmed in asecond experiment, which gave a mean growth delay of 1.0 ±0.9 daysin the absence of radiation but a larger and significant (P < 0.01)growth delay (relative to radiation only) of 10.9 ±4.4 days whencombined with radiation.

DISCUSSION

Hypoxia-selective Toxicity of NNB in Culture. This study demonstrates that, under conditions of prolonged exposure (5-8 h), the

bisnitroimidazole NNB has higher potency and selectivity as a hypoxic cell toxin than simple 2-nitro- or 5-nitroimidazoles. In compar

ison with its mononitro analogue, INB, the bisnitro compound is

CB

ero

0.7 0.5 0.3

3H Retention0.2 1 0.7 0.5 0.3

H RetentionFig. 8. Alkaline elution tests for DNA interstrand cross-links after treatment of hypoxic

AA8 cells with drugs for 5 h. A, NNB: O, •,control; D, • 1 mM; A, A. 1.75 mM; V,V. 2.5 mM. B. chlorambucil: O. •,control; •18 JIM; A, 36 JIM; V, T, 72 (¿M;O. *, 144(¿M.O, G, A, V, O, unirradiated: ».•A. V. ».irradiated (3 Gy).

30 40 0

Time (days)

Fig. 9. Activity of MISO (A) or NNB (B) against hypoxic cells in i.m. MDAH-MCa-4

tumors. Growth curves are for the median mouse in each group of 7 animals, ranked at aleg + tumor diameter of 13 mm. and arc from a single experiment. O, control; D. singledose drug; A, four drug doses (every 3 h for 4 treatments): •,single dose radiation (20Gy); •.single drug dose 30 min before radiation; A. four drug doses (every 3 h for 4treatments) irradiated 30 min after the last dose. All drug injections were ¡.p.at 2.5mmol/kg/dose (0.05 ml/g body weight).

577

Research. on February 13, 2020. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

BIS-NITROIMIDAZOLE-BIOREDUCTIVE DRUGS

50-fold more toxic under hypoxia and 12-fold more selective following 8 h exposure. The alkylating 2-nitroimidazole RB 6145 hashypoxic selectivity comparable to NNB in the repair-defective UV4

cell line, but NNB is the more selective compound in the parentalAA8 cell line (Fig. 4). Thus NNB may have superior hypoxic selectivity against repair-proficient tumor cells, although examination of

differential toxicity in other cell lines is required. The marked sensitivity of hypoxic UV4 cells (or other lines of the same complementation group) to RSU 1069, the active form of RB 6145, has beenreported by others and is considered to reflect the increased contribution of cross-linking to cytotoxicity under hypoxic conditions (29,

45). The present data are consistent with the action of RB 6145 as anRSU 1069 prodrug in this system. RB 6145 is known to be unstablein culture medium with rapid conversion (r,/2 about 30 min at pH 7.4,23°C)to the aziridine RSU 1069 and to the less toxic oxazolodin-2-

one (34). Hence extensive conversion to RSU 1069 would haveoccurred by the time of addition of cells (60 min after addition ofdrug, 37°C)in the present experiments.

Is NNB a Bis-bioreductive Agent? Several lines of evidence areconsistent with the hypothesis that NNB is a bis-bioreductive drug: (a)

its hypoxic potency and selectivity is greater than that of simplemono-2-nitroimidazoles (Figs. 3, 4, and 7), including the close ana

logue INB. This high potency is not accounted for by enhanced celluptake since the uptake factors (intracellular/extracellular concentration) for NNB and INB are identical (0.33) in hypoxic AA8 cultures(33). NNB thus resembles the nitroacridine A/oxides which appear toowe their very high selectivity for hypoxic cells to their bis-bioreduc

tive character (32); (b) the greater potency and selectivity of NNBappear not to be accounted for by interaction between 2- and 5-ni-

troimidazoles which are not linked in the same molecule, as demonstrated using equimolar combinations of MISO and METRO (Fig. 6);(c) concentration-time relationships for cell killing by NNB are dif

ferent from the other nitroimidazoles tested in that the cytotoxicpotency increases markedly with time under hypoxic conditions. Thismay reflect the slow activation of a second reducible center to generate a bifunctional molecule. The cytotoxic potency of NNB at l h issimilar to those of Miso and INB suggesting that initial toxicity is dueto the 2-nitroimidazole moiety and that the slow second activation isdue to reduction of the lower reduction potential 5-nitro group.

The increase in hypoxic potency of NNB with time is not readilyaccounted for by slow consumption of residual oxygen since theincrease in rate of killing is much greater than is observed afterprolonged pregassing; further, the mononitroimidazoles tested [including METRO, which is known to be highly oxygen sensitive inCHO cultures (44)] show a much smaller increase in activity withtime than does NNB. The observation that the CTlt) for NNB isconstant under oxic conditions also implies that the time dependenceunder hypoxia is specifically a consequence of slow metabolicactivation at a second center. This interpretation is supported byrecent experiments showing that after exposure of hypoxic AA8cells to NNB for 2 h followed by extensive washing (removing99.95% of the parent drug), cells continue to die when incubatedunder hypoxic conditions (46). Similar results have been reportedby Stratford et al. (24) for hypoxic exposure of V79 cells afteraerobic treatment with RSU 1069 and were also interpreted asindicating metabolic activation of a covalently bound nitroimida-

zole at the target site.A Locally Doubly Damaged Site Model. If NNB has the biolog

ical characteristics of a bis-bioreductive agent, then the possible

formation of some highly cytotoxic bistranded DNA lesion should beconsidered. We propose a LDDS model analogous with the locallymultiply damaged site model, as advanced by Ward (47, 48) toaccount for the cytotoxicity of ionizing radiation in terms of clustered

DNA lesions. In the LDDS model, bis bioreduction of NNB isproposed to generate a species which is highly cytotoxic because theDNA lesions formed by the two functional domains of the moleculeare closely linked spatially and can thus give rise to damage siteswhich are closely paired on opposite strands of the DNA. Such duplexlesions (e.g., pairs of ssb, a monoadduct or abasic site associated witha ssb, interstrand cross-link) may be highly cytotoxic because they are

difficult to repair with fidelity. This LDDS model is analogous to thatproposed for the cytotoxicity of bleomycin and the enediyne antibiotics, which generate bistranded DNA lesions (49). In the latter casesuch lesions result from a single oxidizing biradical species which canabstract H atoms from deoxyribose sugars in opposite strands (50, 51)to generate dsb in high yield (52, 53).

The lack of ssb formation by NNB in comparison with othernitroimidazoles at equitoxic concentrations (Fig. 7) is suggestiveevidence for the generation of a more toxic duplex DNA lesion. Thecritical cytotoxic lesions do not appear to be DNA dsb analogous tothose formed by ionizing radiation or DNA interstrand cross-links.That cross-links are not involved is strongly suggested by the lack of

hypersensitivity of UV4 cells to NNB under either aerobic or hypoxicconditions (Fig. 3) since this line is sensitive (6-60-fold with respectto AA8) to all DNA cross-linkers tested (43, 45, 54). Alkaline elutionassays in combination with radiation also failed to detect cross-links

in either AA8 or UV4 cells at concentrations well above the cytotoxicrange, while cross-linking by chlorambucil could be readily demonstrated under these conditions (Fig. 8). Reduction products of 2-ni-

troimidazoles can alkylate DNA (55, 56), but the lack of DNAalkylation on metabolic reduction of 5-nitroimidazoles (57) may preclude formation of a cross-link on bisnitroreduction of NNB. However, a variety of other types of nucleoprotein damage by 2- and5-nitroimidazole reduction products might contribute to the formation

of a lethal duplex lesion. Nitro(hetero)aromatic compounds are efficient inducers of DNA ssb (58-60) as illustrated for METRO and

MISO in Fig. 7, and when reduced electrochemically METRO is evenmore effective than MISO in inactivating phage DNA (61). In addition, the nitroso reduction products of 2-nitroimidazoles can alkylateprotein thiol groups (62). Further, reduced 2-nitroimidazoles are

known to bind covalently to tissue macromolecules with retention ofthe N-l side chain (63, 64); thus covalently bound products resultingfrom reduction of the 2-nitro group are expected to retain a 5-nitroi

midazole moiety which might induce further local damage whensubsequently reduced. Therefore duplex lesions such as a DNA monoadduct opposing a ssb, a protein thiol adduci in close association witha ssb, or an intrastrand cross-link might provide the enhancement of

toxicity required for the bisbioreduction/LDDS model.DNA lesion clustering may be important in the action of bioreduc-

tive drugs other than NNB. RB 6145 and RSU 1069 probably owetheir high selectivity and potency to generation of LDDS in the formof DNA cross-links (29, 65). In addition, the high frequency of dsb inChinese hamster ovary cells treated with the benzotriazine-di-A'-oxide

tirapazamine (SR 4233) (66, 67) has been suggested to reflect theproduction of a high local concentration of oxidizing radicals becauseof the relative immobility and physical proximity to DNA of thecritical reductase(s) responsible for SR 4233 activation (67, 68).

Therapeutic Potential of Bisnitroimidazoles. NNB is able toeliminate hypoxic cells in MDAH-MCa-4 mammary tumors as judged

by its ability to extend the tumor growth delay induced by ionizingradiation (Fig. 9). It is similarly active against hypoxic cells in theKHT tumor as assessed by clonogenic assay (33). In both cases,activity was observed only when multiple doses were used, with nosignificant activity after the highest single dose (2.5 mmol/kg) whichcould be administered because of solubility limitations. In contrast,the activity of MISO was not improved by repeated dosing. The latter

578

Research. on February 13, 2020. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

BIS-NITROIMIDAZOLE-BIOREDUCTIVE DRUGS

result was expected since enhancement of radiation response withMISO is due primarily to radiosensitization rather than killing ofhypoxic cells (22), and its half-life in the mouse [1.5 h at 2.5 mmol/kg

(69)] is too short to provide substantial accumulation in the tumorwith three hourly dosing intervals. The improvement in activity ofNNB with multiple doses suggests that activity is probably related tothe concentration-time integral rather than peak concentration in the

tumor at the time of irradiation and is thus likely to be due to hypoxiccell toxicity rather than direct (electron affinic) radiosensitization.However, pharmacokinetic studies are required to confirm this conclusion. The potency of NNB as a hypoxic cell radiosensitizer in vitrois no greater than that of simple mononitroimidazoles (33), but thisproperty may play some part in its activity in vivo. Nitroimidazoleradiosensitizers have generally shown disappointing activity in clinical studies (5), and there are theoretical reasons for expecting thatbioreductive drugs will be superior in radiotherapy (14-16). Further

development of bisnitroimidazoles related to NNB would thereforewarrant a high priority if their antitumor activity in combination withradiation is due primarily to hypoxic cell killing. Recent studies in thislaboratory have identified a bis(2-nitroimidazole) derivative, NSC

661066, with improved water solubility and higher hypoxic potency invitro at early exposure times,4 and which is highly active against KHTtumors when administered after irradiation.5 This postirradiation ac

tivity strongly implicates hypoxic cell killing as the major mechanismand indicates that the bisnitroimidazoles are a novel class ofbioreductive agent which warrant further exploration.

ACKNOWLEDGMENTS

We thank Susan M. Pullen and Maryann L. Taylor for technical assistance,Alan Stewart for radiation dosimetry, Dr. Stephen Cliffe for assistance withoxygen tension measurements, and Drs. Helen B. Stone and Richard P. Hill foradvice with the tumor growth delay experiments.

REFERENCES

1. Chapman, J. D. Measurement of tumor hypoxia by invasive and non-invasive procedures: a review of recent clinical studies. Radiother. Oncol., 20 (Suppl.): 13^19,

1991.2. Henk, J. M., and Smith, C. W. Radiotherapy and hyperbaric oxygen in head and neck

cancer. Lancet, 2: 104-105, 1977.

3. Bush, R. S., Jenkin, R. D. T., All!, W. E. C., Beale, F. A., Bean, H., Dembo, A. J.,and Pringle, J. F. Definitive evidence for hypoxic cells influencing cure in cancertherapy. Br. J. Cancer, 37: 302-306, 1978.

4. Gatenby, R. A., Kessler, H. B., Rosenblum, J. S., Coia, L. R.. Moldofsky, P. J.. andHartz, W. H. Oxygen distribution in squamous cell carcinoma métastasesand itsrelationship to outcome of radiation therapy. Int. J. Radiât.Oncol. Biol. Phys., 14:831-838, 1988.

5. Overgaard, J. Importance of tumour hypoxia in radiotherapy: a meta-analysis of

controlled clinical trials. Radiother. Oncol., 24: S64, 1992.6. Hockel, M., Knoop. C., Schlenger, K., Vorndran, B., Baubmann, E., Mitze, M.,

Knapstein, P. 0., and Vaupel, P. Intratumoral pO2 predicts survival in advancedcancer of the uterine cervix. Radiother. Oncol., 26: 45-50, 1993.

7. McNally, N. J., Denekamp, J., Sheldon, P., Flockhart, I. R., and Stewart, F. A.Radiosensitization by misonidazole (Ro 07-0582). The importance of timing andtumor concentration of sensitizer. Radiât.Res., 73: 568-580, 1978.

8. Denekamp, J., Hirst, D. G., Stewart, F. A., and Terry, N. H. A. Is tumour radiosensitization by misonidazole a general phenomen? Br. J. Cancer, 41: 1-9, 1980.

9. Hall, E. J., and Roizin-Towle, L. Hypoxic sensitizers: radiobiological studies at thecellular level. Radiology, 117: 453-457, 1975.

10. Mohindra, J. K., and Rauth, A. M. Increased cell killing by metronidazole andnitrofurazone of hypoxic compared to aerobic mammalian cells. Cancer Res., 36:930-936, 1976.

11. Moore, B. A., Palcic, B., and Skarsgard, L. D. Radiosensitizing and toxic effects ofthe 2-nitroimidazole Ro-07-0582 in hypoxic mammalian cells. Radiât.Res., 67:459-473, 1976.

12. Brown, J. M., and Lemmon, M. J. Potentiation by the hypoxic cytotoxin SR 4233 of

4 M. P. Hay, H. H. Lee, W. R. Wilson, P. B. Roberts, and W. A. Denny. Hypoxia-

selective antitumor agents. 10. Bis(nitroimidazoles) and related bis(nitroheterocycles):development of derivatives with higher rates of metabolic activation under hypoxia andimproved aqueous solubility, submitted for publication.

5 W. R. Wilson and X. Xu, unpublished observations.

cell killing produced by fractionated irradiation of mouse tumors. Cancer Res., 50:7745-7749, 1990.

13. Adams, G. E. Redox. radiation, and reductive bioactivation. Radial. Res., 132:129-139, 1992.

14. Wilson, W. R. Tumour hypoxia: challenges for cancer chemotherapy. In: M. J.Waring and B. A. J. Ponder (eds.), Cancer Biology and Medicine, Vol. 3, pp. 87-131.

Lancaster, United Kingdom: Kluwer Academic Publishers, 1992.15. Brown, J. M., and Koong, A. Therapeutic advantage of hypoxic cells in tumors: a

theoretical study. J. Nati. Cancer. Inst., 83: 178-185, 1991.16. Denny, W. A., and Wilson, W. R. Bioreducible mustards: a paradigm for hypoxia-

selective prodrugs of diffusible cytotoxins (HPDCs). Cancer Metastasis Rev., 12:135-151, 1993.

17. Brown, J. M., and Giaccia, A. J. Tumour hypoxia: the picture has changed in the1990s. Int. J. Radial. Biol., 65: 95-102, 1994.

18. Chapman, J. D., Franko, A. J., and Sharplin, J. A marker for hypoxic cells in tumourswith potential clinical applicability. Br. J. Cancer, 43: 546-550, 1981.

19. Hodgkiss, R. J., Jones, G., Long, A., Parrick, J., Smith, K. A., Stratford, M. R. L., andWilson, G. D. Flow cytometric evaluation of hypoxic cells in solid experimentaltumours using fluorescence immunodetection. Br. J. Cancer, 63: 119-125, 1991.

20. Raleigh, J. A., Miller, G. G., Franko, A. J., Koch, C. J., Fuciarelli, A. F., and Kelly,D. A. Fluorescence immunohistochemical detection of hypoxic cells in spheroids andtumours. Br. J. Cancer, 56: 395-396, 1987.

21. Koh, W-J., Rasey, J. S., Evans, M. L., Grierson, J. R., Lewellen, T. K., Graham.M. M.. Krohn, K. A., and Griffin. T. W. Imaging of hypoxia in human tumors with[F-lSJfluoromisonidazole. Int. J. Radial. Oncol. Biol. Phys., 22: 199-212, 1991.

22. Rauth, A. M., Paciga, J. E., and Mohindra, J. K. in vivo studies of the cytotoxicity ofhypoxic cell radiosensitizers. In: L. W. Brady (ed.). Radiation Sensitizers: Their Usein the Clinical Management of Cancer, pp. 207-214. New York: Masson Publishing,

1980.23. Adams, G. E., Ahmed, I., Sheldon, P. W.. and Stratford, I. J. Radiation sensitizaron

and chemopotentiation: RSU 1069, a compound more efficient than misonidazole invitro and in vivo. Br. J. Cancer, 49: 571-577, 1984.

24. Stratford, I. J., O'Neill, P., Sheldon, P. W., Silver, A. R. J., Walling, J. M., and

Adams, G. E. RSU 1069, a nitroimidazole compound containing an aziridine group.Bioreduction greatly increases cytotoxicity under hypoxic conditions. Biochem. Phar-macol., 35: 105-109, 1986.

25. Chaplin, D. J., Durand, R. E., Stratford, I. J., and Jenkins, T. C. The radiosensitizingand toxic effects of RSU-1069 on hypoxic cells in a murine tumor. Int. J. Radiât.Oncol. Biol. Phys., 12: 1091-1095, 1986.

26. Hill, R. P., Gulyas, S., and Whitmore, G. F. Studies of the in vivo and in vitrocytotoxicity of the drug RSU-1069. Br. J. Cancer, 53: 743-751, 1986.

27. Cole, S., Stratford, I. J., Fielden, E. M., Adams, G. E., Leopold, W.. Elliott, W., Suto,M., and Sebolt-Leopold, J. Dual function nitroimidazoles less toxic than RSU 1069:selection of candidate drugs for clinical trial (RB 6145 and/or PD 130908). Int.J. Radiât.Oncol. Biol. Phys., 22: 545-548, 1992.

28. Stratford, I. J., Walling, J. M., and Silver, A. R. J. The differential cytotoxicity of RSU1069: cell survival studies indicating interaction with DNA as a possible mode ofaction. Br. J. Cancer, 53: 339-344, 1986.

29. Whitmore, G. F., and Gulyas, S. Studies on the toxicity of RSU-1069. Int. J. Radial.Oncol. Biol. Phys., 12: 1219-1222, 1986.

30. O'Neill, P., McNeill, S. S., and Jenkins, T. C. Induction of DNA crosslinks in vitro

upon reduction of the nitroimidazole-aziridines RSU-1069 and RSU-1131. Biochem.Pharmacol., 36: 1787-1792, 1987.

31. Ahmed, I., Jenkins, T. C., Walling, J. M., Stratford, I. J., Sheldon, P. W.. Adams.G. E., and Fielden, E. M. Analogues of RSU-1069: Radiosensitization and toxicity invitro and in vivo. Int. J. Radiât.Oncol. Biol. Phys., 12: 1079-1081, 1986.

32. Wilson, W. R., van Zijl, P., and Denny, W. A. Bis-bioreductive agents as hypoxia-selective cytotoxins: nitracrine /V-oxide. Int. J. Radial. Oncol. Biol. Phys., 22:693-696, 1992.

33. Hay, M. P., Wilson, W. R., Moselen, J. W., Palmer, B. D., and Denny, W. A.Hypoxia-selective antitumour agents. 8. Bis-(nitroimidazolyl)alkanecarboxamides: anew class of hypoxia-selective cytotoxins and hypoxic cell radiosensitisers. J. Med.Chem., 37: 381-391, 1994.

34. Jenkins, T. C., Naylor, M. A., O'Neill, P., Threadgill, M. D., Cole, S.. Stratford, I. J.,

Adams, G. E., Fielden, M., Suto, M. J., and Stier, M. A. Synthesis and evaluation ofa[[(2-haloethyl)amino]methyl]-2-nitro-l//-imidazole-l-ethanols as prodrugs ofa-[(l-aziridinyl)methyl]-2-nitro-l//-imidazole-l-ethanol (RSU-1069) and its analogues which are radiosensitizers and bioreductively activated cytotoxins. J. Med.Chem., 33: 2603-2610, 1990.

35. Thompson, L. H., Fong, S., and Brookman, K. Validation of conditions for efficientdetection of HPRT and APRT mutations in suspension-cultured Chinese hamsterovary cells. Mutât.Res., 74: 21-36, 1980.

36. Thompson, L. H., Rubin, J. S., Cleaver, J. E., Whitmore, G. F., and Brookman, K. Ascreening method for isolating DNA repair-deficient mutants of CHO cells. SomaticCell Genet., 6: 391-405, 1980.

37. Sum, B. G., Atwell, G. J., and Wilson, W. R. Oxygen dependence of the cytotoxicityand metabolic activation of 4-alkylamino-5-nitroquinoline bioreductive drugs. Br.J. Cancer, 70: 596-603, 1994.

38. Koch, C. J. Polarographic oxygen sensor. U. S. Patent 5030036, 1991.39. Kohn, K. W., Ewigra, G., Erickson. L. C., and Zwelling, L. A. Measurement of strand

breaks and cross-links by alkaline elution. In: E. Friedberg and P. Hanawalt (eds.),DNA Repair. A Laboratory Manual of Research Procedures, pp. 379-401. NewYork: Marcel Dekker, Inc., 1981.

40. Bradley, M. O., and Kohn. K. W. X-ray induced DNA double strand break productionand repair in mammalian cells as measured by neutral filter elution. Nucleic AcidsRes., 7: 793-804, 1979.

579

Research. on February 13, 2020. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

BIS-NITROIMIDAZOLE-BIOREDUCnVE DRUGS

41. Silobrcic, V., and Suit, H. D. Tumor-specific antigen(s) in a spontaneous mammarycarcinoma of C3H mice. I. Quantitative cell transplants into mammary-tumor-agent-positive and-free mice. J. Nati. Cancer Inst., 39: 1113-1119, 1967.

42. Hoy, C. A., Salazar, E. P., and Thompson, L. H. Rapid detection of DNA-damagingagents using repair-deficient CHO cells. Mutât.Res., 130: 321-332, 1984.

43. Hoy, C. A., Thompson, L. H., Mooney, C. L., and Salazar, E. P. Defective DNAcross-link removal in Chinese hamster cell mutants hypersensitive to bifunctionalalkylating agents. Cancer Res., 45: 1737-1743, 1985.

44. Rauth, A. M., McClelland, R. A., Michaels. H. B., and Battistella, R. The oxygendependence of the reduction of nitroimidazoles in a radiolytic model system. Int.J. Radiât.Oncol. Biol. Phys., 10: 1323-1326, 1984.

45. Whitmore, G. F., Dulhanty, A., Varghese, A. J., and Gulyas, S. Response of repairand reduction deficient mutants to agents requiring bioreduction. In: G. E. Adams etal. (eds.). Selective Activation of Drugs by Redox Processes, pp. 285-294. New York:Plenum Publishing Corp., 1990.

46. Wilson, W. R., Moselen, J. W., Hay, M. P., and Denny, W. A. Bis-nitroimidazolebioreductive drugs: evidence for in situ reduction of DNA-bound metabolites. Abstract P35-588. In: Forty-second Annual Meeting, Radiation Research Society, 1994.

47. Ward, J. F. Symposium on Radical Processes in Radiobiology and Carcinogenesis.Radiât.Res., 86: 185-195, 1981.

48. Ward, J. F. Biochemistry of DNA lesions. Radiât.Res., 104: S103-S111, 1985.49. Dedon, P. C., and Goldberg, I. H. Free radical mechanisms involved in the formation

of sequence-dependent bistranded DNA lesions by the antitumor antibiotics bleomy-cin, neocarzinostatin and calicheamicin. Chem. Res. Toxicol., 5: 311-332, 1992.

50. Zein, N., McGahren, W. J., Morton, G. O., Ashcroft, J., and Ellestad, G. A. Exclusiveabstraction of non-exchangeable hydrogens from DNA by calicheamicin yl'. ]. Am.

Chem. Soc., ///: 6888-6890, 1989.51. Meschwitz, S. M., and Goldberg, I. H. Selective abstraction of 2H from C5' of

thymidylate in oligodeoxynucleotide by radical center at C6 of diradical species ofneocarzinostatin: chemical evidence for structure of activated drug-DNA complex.Proc. Nati. Acad. Sci. USA, 88: 3047-3051, 1991.

52. Hatayama, T., and Goldberg, I. H. DNA damage and repair in relation to cell killingin neocarzinostatin-treated HeLa cells. Biochim. Biophys. Acta, 563: 59-71, 1979.

53. Shiloh, Y., van der Schans, G. P., Lohman, P. H. M., and Becker, Y. Induction andrepair of DNA damage in normal and ataxia telangiectasia skin fibroblasts treatedwith neocarzinostatin. Carcinogenesis (Lond.), 4: 917—921,1983.

54. Palmer, B. D., Wilson, W. R., Cliffe, S., and Denny, W. A. Hypoxia-selectiveantitumor agents. 5. Synthesis of water-soluble nitroaniline mustards with selectivecytotoxicity for hypoxic mammalian cells. J. Med. Chem., 35: 3214-3222, 1992.

55. Varghese, A. J., and Whitmore, G. F. Binding to cellular macromolecules as a

possible mechanism for the cytotoxicity of misonidazole. Cancer Res., 40:2165-2169, 1980.

56. Miller, G. G., Ngan-Lee, J., and Chapman, J. D. Intracellular localization of radio-actively labeled misonidazole in EMT-6 tumour cells in vitro. Int. J. Radiât.Onco!.Biol. Phys., 8: 741-744, 1982.

57. Malliaros, D. P., and Goldman, P. Interaction of metronidazole with Escherichia colideoxyribonucleic acid. Biochem. Pharmacol., 42: 1739—1744, 1991.

58. Palcic, B., and Skarsgard, L. D. Cytotoxicity of misonidazole and DNA damage inhypoxic mammalian cells. Br. J. Cancer, 37: 54-59, 1978.

59. Murray, D., and Meyn, R. E. DNA damage in normal and neoplastic mouse tissuesafter treatment with misonidazole in vivo. Biochem. Pharmacol., 34: 3275-3279,1985.

60. Olive, P. L. Mechanisms of the in vitro toxicity of nitroheterocycles including Flagyland misonidazole. In: L. W. Brady (ed.), Radiation Sensitizers. Their Use in theClinical Management of Cancer, pp. 39-44. New York: Masson Publishing, 1980.

61. Lafleur, M. V. M., Zahoor, A., Knight, R. C., Pluijmackers-Westmijze, E. J., Loman,H., and Edwards, D. I. Electrolytic reduction of nitroheterocyclic drugs leads tobiologically important damage in DNA. Int. J. Radiât.Biol., 48: 917-925, 1985.

62. Beruhe, L. R., Farah, S., McClelland, R. A., and Rauth, A. M. Effect of 1-methyl-2-nitrosoimidazole on intracellular thiols and calcium levels in Chinese hamster ovarycells. Biochem. Pharmacol., 42: 2153-2161, 1991.

63. Chapman, J. D., Baer, K., and Lee, J. Characteristics of the metabolism-inducedbinding of misonidazole to hypoxic mammalian cells. Cancer Res., 43: 1523-1528,

1983.64. Franko, A. J., and Garrecht, B. M. Misonidazole retention by normal tissues: a

distinction between label on the ring and side chain. Int. J. Radiât.Oncol. Biol. Phys.,12: 1259-1262, 1986.

65. O'Neill, P., McNeil, S. S., and Jenkins, T. C. Induction of DNA crosslinks in vitro

upon reduction of the nitroimidazole-aziridines RSU-1069 and RSU-1131. Biochem.Pharmacol., 36: 1787-1792, 1987.

66. Zeman, E. M., and Brown, J. M. Pre- and post-irradiation radiosensitization by SR4233. Int. J. Radiât.Oncol. Biol. Phys., 16: 967-971, 1989.

67. Wang, J., Beidermann, K. A., and Brown, J. M. Repair of DNA and chromosomebreaks in cells exposed to SR 4233 under hypoxia or to ionizing radiation. CancerRes., 52: 4473-4477, 1992.

68. Brown, J. M. SR 4233 (tirapazamine): a new anticancer drug exploiting hypoxia insolid tumours. Br. J. Cancer, 67: 1163-1170, 1993.

69. Workman, P. Pharmacokinetics of hypoxic cell radiosensitizers. A review. In: L. W.Brady (ed.), Radiation Sensitizers. Their Use in the Clinical Management of Cancer,pp. 192-206. New York: Masson Publishing, 1980.

580

Research. on February 13, 2020. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

1995;55:574-580. Cancer Res   John W. Moselen, Michael P. Hay, William A. Denny, et al.   Selectivity as a Bioreductive Drugmide (NSC 639862), a Bisnitroimidazole with Enhanced-[2-(2-Methyl-5-nitroimidazolyl)ethyl]-4-(2-nitroimidazolyl)butanaN

  Updated version

  http://cancerres.aacrjournals.org/content/55/3/574

Access the most recent version of this article at:

   

   

   

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

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

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

.http://cancerres.aacrjournals.org/content/55/3/574To request permission to re-use all or part of this article, use this link

Research. on February 13, 2020. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from