Studies on the mechanisms of the radiosensitizing and cytotoxic properties of RSU-1069 and its analogues

Nitroimidazoles as hypoxic cell radiosensitizers and hypoxia probes: misonidazole, myths and mistakes

Nitroimidazoles have been extensively explored as hypoxic cell radiosensitizers but have had limited clinical success, with efficacy restricted by toxicity. However, they have proven clinically useful as probes for tumour hypoxia. Both applications, and probably much of the dose-limiting toxicities, reflect the dominant chemical property of electron affinity or ease of reduction, associated with the nitro substituent in an aromatic structure. This single dominant property affords unusual, indeed extraordinary flexibility in drug or probe design, suggesting further development is possible in spite of earlier limitations, in particular building on the benefit of hindsight and an appreciation of errors made in earlier studies. The most notable errors were: the delay in viewing cellular thiol depletion as a likely common artefact in testing in vitro; slow recognition of pH-driven concentration gradients when compounds were weak acids and bases; and a failure to explore the possible involvement of pH and ascorbate in influencing hypoxia probe binding. The experience points to the need to involve a wider range of expertise than that historically involved in many laboratories when studying the effects of chemicals on radiation response or using diagnostic probes.

Nitro-imidazole radiosensitizer-induced toxicity in cynomolgus monkeys

Intravenously administered nitro-imidazole radiosensitizer and alkylating anticancer compound CI-1010, designated as (R)-alpha-[[(2-bromoethyl)amino]methyl]-2-nitro-1H-imidazole-1-ethanol monohydrobromide, causes multiorgan toxicity in rodents, including retinal degeneration. This study determined the potential of CI-1010 to induce similar effects in nonhuman primates. One male and 1 female cynomolgus monkey were given single daily doses of CI-1010 intravenously for 5 consecutive days each week for 3 wk. Doses were escalated from 5 mg per kilogram of body weight in week 1 to 40 and 60 mg/kg in week 3. Postdosing emesis occurred in both monkeys at 5 mg/kg, and clinical signs at 40 and 60 mg/kg included more pronounced emesis, reduced food consumption, pallor, weakness, and body weight loss. At study termination, both monkeys had markedly reduced peripheral blood lymphocytes and moderately lowered erythrocyte, hemoglobin, and hematocrit levels, which correlate with a decreased total nucleated bone marrow cell count. At necropsy, the monkeys had pancytic bone marrow hypocellularity, multiorgan lymphoid depletion, pancreatic acinar cell apoptosis, testicular seminiferous tubular degeneration, and bilateral multifocal retinal degeneration involving the photoreceptor and outer nuclear layers. Ultrastructurally, selected inner and outer retinal rod segments were swollen and fragmented, a state associated with cytoplasmic condensation and pyknosis of the outer nuclear cell layer. Thus, CI-1010 induced toxicity of hematopoietic/lymphoid organs, retina, testes, and pancreas in monkeys, findings similar to those of previous studies in rodents.

Retinal degeneration in rats induced by CI-1010, a 2-nitroimidazole radiosensitizer

The anti-cancer compound CI-1010, designated as (R)-alpha-([(2-bromoethyl)amino]methyl)-2-nitro-1H-imidazole-1-ethanol monohydrobromide, has a proposed dual mechanism of action due to alkylating and radiosensitizing activities. To assess potential toxicity, adult Wistar rats were treated with a single intravenous injection (0, 50, 100, 150, 225, or 350 mg/kg) and necropsied at 4 or 29 days following treatment. In a repeated dose experiment, rats were injected daily (0, 10, 40, or 80 mg/kg; 5 doses/wk) for 3 wk and necropsied at the end of week 3 or 7. CI-1010 induced retinal degeneration by 4 days after a single injection of > or = 225 mg/kg or by 3 wk of repeated injections of > or = 40 mg/kg. The locally extensive to diffuse retinal degeneration involved the photoreceptor and outer nuclear layer. The photoreceptor layer was vacuolated and compressed corresponding to ultrastructural evidence of inner segment swelling and outer segment fragmentation. The outer nuclear layer was thinned due to loss of nuclei and contained numerous pyknotic or karyorrhectic nuclei. These nuclear changes were morphologically consistent with apoptosis and many outer nuclear layer nuclei labeled with in situ TdT-mediated dUTP-digoxigenin nick end labeling (Apoptag). The retinal degeneration was nonreversible, evidenced by increased lesion severity and incidence after CI-1010 was withdrawn for either 25 or 28 days.

Mechanism of oncogenicity for bioreductive drugs

The oncogenic transforming potential of a series of bifunctional bioreductive drugs were examined under either aerated or hypoxic conditions to assess the contribution of side chains or nitroreduced products toward their carcinogenic mechanisms. Both the cytotoxicity and transforming effects of these drugs increased as a function of dose under hypoxia. In air and at doses that resulted in comparable cell killing, RSU-1069 and RB-88716 were substantially more oncogenic than RSU-1164 or SR-4233. In nitrogen, the oncogenicity of SR-4233 as a function of survival increased, whereas the transforming effect for the aziridine-containing drugs, RSU-1969 and RB-88716, decreased. These data suggest that, among the drugs examined, the transforming moiety in air is largely a function of the alkylating aziridine group. In hypoxia, the reduction of the nitro-moiety to the corresponding active metabolites may be responsible for much of the transformation observed.

Assessment of a range of novel nitro-aromatic radiosensitizers and bioreductive drugs

In a directed search for the best compounds for clinical evaluation, some 150 selected nitroaromatic compounds, representing 6 distinct types, namely, furans, thiophenes, imidazoles, pyrazoles, pyrroles, and triazoles, have been synthesized and tested as hypoxic cell radiosensitisers and bioreductive drugs. These compounds have a wide range of one-electron redox potentials, ranging from -700 mV for 3-nitropyrroles to -250 mV for 5-nitrofurans. Within each series, those agents bearing alkylating moieties on the side chain are generally the more effective radiosensitisers in vitro. Studies in vivo demonstrated that the bifunctional nitroimidazoles were superior to the other nitroarenes tested. In terms of bioreductive cell killing, the best differential between oxic and hypoxic cell toxicity was shown for the bifunctional 2-nitroimidazoles, which had values greater than 20. In contrast, the other classes of nitroarines generally showed differential toxicities of less than 10.

The repair of DNA damage induced in V79 mammalian cells by the nitroimidazole-aziridine, RSU-1069. Implications for radiosensitization

The induction and repair of single (ssb) and double (dsb) strand breaks in DNA under aerobic or hypoxic conditions have been determined using sucrose sedimentation techniques following incubation of V79 mammalian cells with RSU-1069 or misonidazole, representative of a conventional 2-nitroimidazole radiosensitizer, for 1-1.5 hr at either 293 or 277 degrees K and subsequent irradiation at 277 degrees K. In all cases, the dose dependences for the induction of strand breaks are linear and consistent with an enhancement in the yield of DNA damage induced by the 2-nitroimidazoles under hypoxic conditions. With RSU-1069 at 293 degrees K, the dose dependence of ssb is displaced reflecting DNA damage induced during pre-incubation. From these dependences, it is evident that the enhanced radiosensitization by RSU-1069 may not be accounted for in terms of accumulation of the agent at DNA. From the repair studies, DNA breaks induced by RSU-1069 in the absence of radiation have been shown to persist for at least 3 hr. With a combination of RSU-1069 and radiation under hypoxic conditions, the repair timescale of the induced breaks is significantly longer and an increase in the residual yields of both ssb and dsb (at 2-3 hr) was observed when compared with the observation in the presence of misonidazole or oxygen. From these studies, it is inferred that the enhanced radiosensitization of RSU-1069 at 293 degrees K is a consequence of the formation of non-repairable DNA damage together with a modification of the repairability of the radiation-induced DNA breaks.

A comparison of the techniques of alkaline filter elution and alkaline sucrose sedimentation used to assess DNA damage induced by 2-nitroimidazoles

The induction of DNA single-strand breaks (DNA-SSB) in Chinese hamster V79-379A lung fibroblasts by misonidazole or RSU-1069 under both aerobic and hypoxic conditions was examined following incubations for up to 4 hr at 310 degrees K using the technique of alkaline filter elution. Incubation with RSU-1069 induces DNA-SSB under both hypoxic and aerobic conditions, whereas incubation with misonidazole induces DNA-SSB only under hypoxia. The yield of breaks is dependent on both agent concentration and contact time. Following identical treatments with these agents, the yield of DNA-SSB (expressed in radiation dose equivalents) determined by alkaline filter elution is about one order of magnitude less than that previously determined by alkaline sucrose gradient sedimentation. In contrast to radiation induced DNA-SSB, alkaline elution is less sensitive than alkaline sucrose gradient sedimentation when determining DNA-SSB induced by RSU-1069 and misonidazole. During the filter elution assay, either increasing cell lysis from 2 to 4 hr, the pH of the lysing buffer from pH 8.7 to 12.5 or the elution buffer from pH 12.2 to 12.5 does not significantly effect the yield of DNA-SSB. Increasing the pH of the lysing or elution buffers to greater than pH 13 however results in considerable degradation of the DNA, whereby 50-85% of the total DNA passes through the filter with the lysing solution. This effect was similar for DNA from both control and chemically insulted cells. In conclusion, it is apparent that incubation with these agents results in the induction of DNA damage which is expressed as a DNA-SSB only after prolonged treatment under alkaline conditions. Further, the use of alkaline elution to study DNA-SSB damage induced chemically must be treated with caution in the light of these findings.

High uptake of RSU 1069 and its analogues melanotic melanomas

RSU 1069 and RSU 1164 are electron affinic agents that contain a nitro group together with a weakly basic alkylating aziridine moiety, and they represent lead compounds in the development of dual-function, bioreductive, hypoxic cell radiosensitizers. We studied the pharmacokinetics of these drugs in mice carrying KHT sarcoma. Lewis lung carcinoma, and B16 melanoma. Following an i.p. dose of 80 mg/kg, absorption was rapid and the elimination t1/2 was in the region of 30 min for both agents. Maximal tumour levels were 91, 16 and 19 microgram/ml for RSU 1069 and 109, 26 and 28 microgram/ml for RSU 1164 in. the B16, KHT and Lewis lung tumours, respectively. In B16 melanoma these levels corresponded to tumour:plasma ratios of 3.8 for RSU 1069 and 3.7 for RSU 1164. Cellular uptake of RSU 1069, RSU 1164 and a related compound, RB 7040, was measured in vitro as a function of extracellular pH. Melanotic cells from both B16 melanoma and HX118, a human tumour xenograft, showed substantially greater accumulation of these weakly basic sensitizers than any other cell type examined. Ratios of intra-:extracellular concentration (Ci/Ce) for RSU 1069 were around unity and independent of pH for Lewis lung cells and HX34 amelanotic melanoma cells, whereas ratios of up to 3 and 5 were obtained in B16 and HX118 cells, respectively. The highest measured value of Ci/Ce was 15 for RSU 1164 in HX118 cells at pH 8.4; this compares with a ratio of 1.5 for HX34 cells at the same pH. These studies indicate that the high levels of uptake of the weakly basic sensitizers into melanotic melanoma in vivo is a cell-mediated phenomenon and may be due to a lower average intracellular pH in the melanotic cells.

Assessment of the repair and damage of DNA induced by parent and reduced RSU-1069, a 2-nitroimidazole-aziridine

The cellular repair and damage of DNA induced by parent and reduced RSU-1069, a 2-nitroimidazole-aziridine, was assessed at both the molecular and cellular level. At the molecular level, after in vitro incubation with parent or reduced RSU-1069, plasmid DNA was transfected into Escherichia coli (AB1157) with subsequent selection for gene expression. For equivalent levels of DNA strand breakage following such treatment it is evident from the relative transformation frequencies that interactions with reduced RSU-1069 lead to DNA damage consistent with bifunctional action of a metabolite(s). At the cellular level, the cytoxicity of RSU-1069 was determined for a series of repair deficient mutants of E. coli under both aerobic and hypoxic conditions. The differential aerobic:hypoxic cytotoxicity ratio is approximately 3. We conclude that the repair of cellular DNA damage induced by RSU-1069 involves activation of the gene products under the control of the recA gene and not those under the control of the ada gene. The ability of cellular systems to repair damage induced by RSU-1069 may play a significant role in determining its efficiency to act as a hypoxic cell radiosensitizer and a hypoxia selective cytotoxin.

A probe for intracellular concentrations of drugs: delayed fluorescence from acridine orange

The aim of this work is to develop fluorescent probes that will indicate effective concentrations of therapeutic agents, or endogenous protectors, at important cellular sites. Acridine orange associates with nucleic acids and emits a 'delayed' fluorescence signal. This signal is quenched by oxidants such as oxygen, nitroaryl radiosensitizers, adriamycin and mitomycin-c, and reductants such as thiols, ascorbate and other radioprotectors. The quenching of the acridine orange delayed fluorescence reflects the effective concentration of these therapeutically-important oxidants and reductants near DNA. The relative concentration of basic radiosensitizers such as pimonidazole (Ro 03-8799) near the DNA is greater than that of misonidazole. Thiols quench the delayed fluorescence signal according to the degree of ionization of the thiol function; this may model the reactivity of thiols with guanine radical sites in DNA. Ascorbate and aminopyrine do not quench the delayed fluorescence from cells stained with acridine orange as these compounds are taken up by cells very inefficiently.

Keynote address: hypoxic cell sensitizers: clinical developments

"Tumours having small islands of cells should be radiosensitive, and those having large masses, radioresistant. This was found to be the case and there is a ready explanation for this: the outside cells are close to blood vessels and will thus be well supplied with oxygen, whilst the inside cells are not well placed in this respect. Now it is known that cells well supplied with oxygen are radiosensitive, compared to the same cells deprived of oxygen; it is therefore likely that for this reason, variation in blood supply, that the outside cells are more easily destroyed than those within."

NLP-1: a DNA intercalating hypoxic cell radiosensitizer and cytotoxin

The 2-nitroimidazole linked phenanthridine, NLP-1 (5-[3-(2-nitro-1-imidazoyl)-propyl]-phenanthridinium bromide), was synthesized with the rationale of targeting the nitroimidazole to DNA via the phenanthridine ring. The drug is soluble in aqueous solution (greater than 25 mM) and stable at room temperature. It binds to DNA with a binding constant 1/30 that of ethidium bromide. At a concentration of 0.5 mM, NLP-1 is 8 times more toxic to hypoxic than aerobic cells at 37 degrees C. This concentration is 40 times less than the concentration of misonidazole, a non-intercalating 2-nitroimidazole, required for the same degree of hypoxic cell toxicity. The toxicity of NLP-1 is reduced at least 10-fold at 0 degrees C. Its ability to radiosensitize hypoxic cells is similar to misonidazole at 0 degrees C. Thus the putative targeting of the 2-nitroimidazole, NLP-1, to DNA, via its phenanthridine group, enhances its hypoxic toxicity, but not its radiosensitizing ability under the present test conditions. NLP-1 represents a lead compound for intercalating 2-nitroimidazoles with selective toxicity for hypoxic cells.

Enhancement of DNA damage in mammalian cells upon bioreduction of the nitroimidazole-aziridines RSU-1069 and RSU-1131

The induction of DNA double-(dsb) and single-(ssb) strand breaks by RSU-1069, RSU-1131 and misonidazole in V79 mammalian cells has been investigated using sedimentation in isokinetic sucrose gradients after incubation for various times (1-3 hr) at 310 K under both hypoxic and aerobic conditions. Double strand breaks are produced by RSU-1069 and RSU-1131 predominantly under hypoxic conditions. Comparison of the cellular DNA damage induced by these agents leads to the following facts: (1) the yield of ssb induced by these agents is substantially increased under hypoxia, (2) RSU-1069 and RSU-1131 are much more effective than misonidazole, on a concentration basis, at causing strand breakage both under hypoxic and aerobic conditions; and (3) RSU-1069 is more efficient on a concentration basis than RSU-1131 at inducing both ssb and dsb under both conditions. From these findings and molecular studies it is suggested that these 2-nitroimidazole aziridines act as monofunctional alkylating agents under aerobic conditions, a factor that governs their aerobic cytotoxicity. Under hypoxic conditions, it is suggested that the induction of dsb and crosslinks by these agents (bifunctional character) may play a major role in determining the ability of such agents to act as hypoxia-selective cytotoxins.

Pharmacokinetics and metabolism of the mixed-function hypoxic cell sensitizer prototype RSU 1069 in mice

RSU 1069 is a leading compound in the class of mixed-function hypoxic cell sensitizers. Possessing an alkylating aziridine function as well as a nitro group, it represents an important prototype molecule for new sensitizer development. Using a novel HPLC assay for RSU 1069 and its metabolites with a cyanopropyl column, we studied the detailed pharmacokinetics and metabolism of this drug in mice. An i.v. dose of 100 mg kg-1 produced peak plasma concentrations of about 100 micrograms ml-1. Absorption was rapid after i.p. injection but peak plasma concentrations were some three- to fourfold lower, giving an i.p. bioavailability of 55%. The elimination t1/2 was route-dependent; e.g. after 50 mg kg-1 the t1/2 was 37.2 and 22.4 min for the i.v. and i.p. routes respectively (P less than 0.001). There was also an indication of dose-dependent kinetics, with a 37% increase in elimination t1/2 when the i.p. dose was doubled from 50 to 100 mg kg-1. Oral bioavailability was low. The volume of distribution was 0.65-1.31 ml g-1 at 50 mg kg-1, but tissue penetration was limited. Brain/plasma ratios ranged from 9.3% to 66.8%, while the mean steady-state tumour/plasma ratio was 28.4%, a value considerably less than the 80%-100% ratios occurring with the neutral 2-nitroimidazole misonidazole. About 18% and 8% of a dose were excreted as the parent drug and the ring-opened hydrolysis product (RSU 1137) in the 8 h urine, indicating the likelihood of extensive metabolism via aziridine-ring removal and nitroreduction. RSU 1137 was also detected in mouse plasma and tissues and, in contrast to the aziridine ring-intact parent compound, gave tumour/plasma ratios of 100%. These studies should provide a pharmacokinetic basis for the evaluation and development of improved mixed-function sensitizers.

Dual-function radiation sensitizers and bioreductive drugs: factors affecting cellular uptake and sensitizing efficiency in analogues of RSU 1069

Alkyl aziridine analogues of the hypoxic cell radiosensitizer RSU 1069 have been synthesized and one of these, RB 7040, containing the tetramethyl substituted aziridine, is a more efficient sensitizer in vitro than RSU 1069 (Ahmed et al., 1986). The extent to which variation in drug uptake can influence the sensitizing efficiency of RSU 1069 and its analogues has been investigated by determining the cellular uptake of these weakly basic sensitizers as a function of the pH of the extracellular medium (pHe) over the range 5.4-8.4. Following exposure of V79 cells to these agents for 1 h at room temperature, the ratio of intra- to extracellular concentration (Ci/Ce) was near unity at pH 5.4. Increasing pHe to 8.4 resulted in no change in the ratio Ci/Ce for RSU 1069 (pKa = 6.04). In contrast, the values of Ci/Ce increased three-fold for RSU 1165 (pKa = 7.38) and eleven-fold for RB 7040 (pKa = 8.45). Radiosensitization by RSU 1069 showed little dependence on pHe over the range studied, whereas increasing pH caused an apparent increase in sensitizing efficiency of both RSU 1165 and RB 7040. However, when the enhancement ratios for sensitization were normalized to take account of the effect of extracellular pH on drug uptake, efficiency of sensitization was independent of pHe. This study suggests that changes in basicity (pKa) may have wider potential for therapeutic exploitation on the basis of selective tumour uptake for this type of agent.

Effects of substituted 2-nitroimidazoles and related compounds on unscheduled DNA synthesis in rat hepatocytes and in non-transformed (BL8) and transformed (JB1) rat liver epithelial derived cell lines

1. Using unscheduled DNA synthesis as an index, the possible interaction of a number of substituted nitroimidazoles, e.g. misonidazole, with cellular DNA has been investigated. Transformed (JB1), non-transformed (BL8) rat liver epithelial derived cell lines and freshly prepared rat hepatocytes have been used. 2. Under anaerobic or aerobic conditions, relative to cells exposed to a nitroquinoline-N-oxide standard, misonidazole and related nitroimidazoles were very poor at stimulating unscheduled DNA synthesis in JB1 or BL8 cells or in hepatocytes, even at the highest concentrations tested (10 mM). Under anaerobic conditions, metabolic activation did occur as judged from the time-dependent depletion of cellular reduced glutathione in all three cell types. 3. It was concluded that in hypoxic cells an important mode of action of such nitroimidazoles as chemotherapeutic sensitisers may be by their interaction with cellular thiols rather from their interaction with DNA. 4. Functionalisation of the nitroimidazole ring with a side chain containing an aziridine function, e.g. RSU-1069 (1-(2-nitro-1-imidazolyl)-3-(1-aziridinyl)-2-propanol), results in the induction of unscheduled DNA synthesis in cells exposed under both aerobic and anaerobic conditions. On a molar basis, however, this induction was not so great as that caused by the simple monofunctional alkylating agent 1-aziridineethanol itself. Methyl-substitution of the aziridine ring in RSU-1069 reduced the extent of unscheduled DNA synthesis. 5. With all the compounds tested, unscheduled DNA synthesis was greater in JB1 cells than in BL8s or in hepatocytes.

Variation of the radiosensitizing efficiency of RSU-1069 with pre-irradiation contact times: a rapid mix study

Using a cellular fast-mixing technique, the time course of radiation sensitization of hypoxic, V79 cells by various concentrations of RSU-1069 (0.25-2.5 mmol dm-3) and misonidazole (2.5-50 mmol dm-3) have been studied to distinguish between fast chemical processes and the much slower biochemical responses to ionizing radiation and the monofunctional alkylating action of RSU-1069. Under conditions of equi-concentration, misonidazole and RSU-1069 show similar radiosensitizing efficiencies for pre-irradiation contact times up to 1 s. The values of the sensitizer enhancement ratio of approximately 1.5 for both 2-nitroimidazoles (2.5 mmol dm-3) is considerably less than that of 1.9-2.8 determined with misonidazole for a pre-irradiation contact time of 1 h under hypoxia. It is proposed that the enhanced radiosensitizing efficiency of RSU-1069 compared to that of misonidazole after long contact times involves, in part, the formation of 'sub-toxic' damage probably involving monofunctional and/or bifunctional action of RSU-1069 prior to irradiation.

Analogues of RSU-1069: radiosensitization and toxicity in vitro and in vivo

A series of nitroimidazoles containing aziridine and alkyl-substituted aziridine functions has been synthesized. The 2-nitroimidazole compounds examined all show greater radiosensitizing efficiency in vitro than misonidazole. The 4- and 5-nitroimidazole analogues are also more efficient than equivalent compounds which do not contain the alkylating aziridine moiety. All the compounds show increased toxicity towards hypoxic cells relative to aerobic cells, but this toxicity is reduced by alkyl-substitution of the aziridine ring. In vivo toxicity can also be reduced by modification of the aziridine function, but such alterations appear to have less effect upon the high radiosensitizing efficiency of these compounds in vivo. As a consequence of this study, compounds of potentially improved therapeutic utility compared to RSU-1069, the first reported member of this class, have been identified.

Experimental pharmacokinetics of RSU-1069 and its analogues: high tumor/plasma ratios

The mixed-function radiosensitizer RSU-1069 and its analogues possess both alkylating and electron-affinic properties, and have been shown to be more efficient radiosensitizers than misonidazole both in vivo and in vitro. The pharmacokinetics following intraperitoneal injection of three members of this series, RSU-1069, RSU-1164, and RSU-1172 have been studied in C57BL mice bearing B16 melanoma. Peak tumor levels of each compound, and tumor/plasma ratios (T/P) were found to be high: T/P RSU-1069 = 3.8, RSU-1164 = 3.7; RSU-1172 = 4.0. In contrast, other normal tissues including brain showed tissue/plasma ratios close to 1. The mechanisms responsible for differential tumor uptake are unknown, but may depend on the basicity of the compounds, leading to preferential uptake in areas of low pH, or alternatively they may be associated with the alkylating function. This group of compounds appear to demonstrate highly favorable tissue distributions.