CYP3A4-dependent cellular response does not relate to CYP3A4-catalysed metabolites of C-1748 and C-1305 acridine antitumor agents in HepG2 cells.

High CYP3A4 expression sensitizes tumor cells to certain antitumor agents while for others it can lower their therapeutic efficacy. We have elucidated the influence of CYP3A4 overexpression on the cellular response induced by antitumor acridine derivatives, C-1305 and C-1748, in two hepatocellular carcinoma (HepG2) cell lines, Hep3A4 stably transfected with CYP3A4 isoenzyme, and HepC34 expressing empty vector. The compounds were selected considering their different chemical structures and different metabolic pathways seen earlier in human and rat liver microsomes C-1748 was transformed to several metabolites at a higher rate in Hep3A4 than in HepC34 cells. In contrast, C-1305 metabolism in Hep3A4 cells was unchanged compared to HepC34 cells, with each cell line producing a single metabolite of comparable concentration. C-1748 resulted in a progressive appearance of sub-G1 population to its high level in both cell lines. In turn, the sub-G1 fraction was dominated in CYP3A4-overexpressing cells following C-1305 exposure. Both compounds induced necrosis and to a lesser extent apoptosis, which were more pronounced in Hep3A4 than in wild-type cells. In conclusion, CYP3A4-overexpressing cells produce higher levels of C-1748 metabolites, but they do not affect the cellular responses to the drug. Conversely, cellular response was modulated following C-1305 treatment in CYP3A4-overexpressing cells, although metabolism of this drug was unaltered.

Diminished toxicity of C-1748, 4-methyl-9-hydroxyethylamino-1-nitroacridine, compared with its demethyl analog, C-857, corresponds to its resistance to metabolism in HepG2 cells.

The narrow "therapeutic window" of anti-tumour therapy may be the result of drug metabolism leading to the activation or detoxification of antitumour agents. The aim of this work is to examine (i) whether the diminished toxicity of a potent antitumour drug, C-1748, 9-(2'-hydroxyethylamino)-4-methyl-1-nitroacridine, compared with its 4-demethyl analogue, C-857, results from the differences between the metabolic pathways for the two compounds and (ii) the impact of reducing and/or hypoxic conditions on studied metabolism. We investigated the metabolites of C-1748 and C-857 formed in rat and human liver microsomes, with human P450 reductase (POR) and in HepG2 cells under normoxia and hypoxia. The elimination rate of C-1748 from POR knockout mice (HRN) was also evaluated. Three products, 1-amino-9-hydroxyethylaminoacridine, 1-aminoacridinone and a compound with an additional 6-membered ring, were identified for C-1748 and C-857 in all studied metabolic systems. The new metabolite was found in HepG2 cells. We showed that metabolic rate and the reactivity of metabolites of C-1748 were considerably lower than those of C-857, in all investigated metabolic models. Compared with metabolism under normoxia, cellular metabolism under hypoxia led to higher levels of 1-aminoacridine and aza-acridine derivatives of both compounds and of the 6-membered ring metabolite of C-1748. In conclusion, the crucial role of hypoxic conditions and the direct involvement of POR in the metabolism of both compounds were demonstrated. Compared with C-857, the low reactivity of C-1748 and the stability of its metabolites are postulated to contribute significantly to the diminished toxicity of this compound observed in animals.

Synthesis and antiproliferative activity of substituted benzopyranoisoindoles: a new class of cytotoxic compounds.

A series of novel aminosubstituted benzopyranoisoindoles possessing structural analogy to an active nitracrine metabolite are reported. The compounds exhibited interesting cytotoxic activity against a panel of cell lines, which was maximized by the presence of both 1-dialkylaminoethyl and 3-nitro substituents.

Products of metabolic activation of the antitumor drug ledakrin (nitracrine) in vitro.

The aim of this work was to characterize the products of metabolic activation of the antitumor drug ledakrin (Nitracrine) in model metabolic systems, where formation of drug-DNA adducts was previously discovered. The metabolic products obtained in different biological systems were compared with those obtained in experiments where chemical reducing agents were applied. Therefore, activation products were obtained in the presence of the microsomal fraction of rat liver and in the experiments with the reducing agents dithiothreitol, hydrazine hydrate, and SnCl(2). Furthermore, transformations of the drug with oxidoreductase enzymes DT-diaphorase and xanthine oxidase were observed. The ledakrin transformation products were separated and analyzed by HPLC with diode array detection. Structural studies of the products were performed by means of ESI-MS and NMR. Proton, carbon, and nitrogen assignments were made based upon DQF-COSY, ROESY, TOCSY, HSQC, and HMBC experiments. It was demonstrated during the reduction of ledakrin that a key metabolite, a compound with an additional five-membered ring attached to positions 1 and 9 of the acridine core and with the retained 9-aminoalkyl side chain, was formed in all the systems that were studied. It was determined that the reactive nitrogen atoms of this additional ring underwent further transformations resulting in the formation of a six-membered ring produced by the addition of a carbon atom to the dihydropyrazoloacridine ring. Furthermore, it was observed that positions 2 and 4 of ledakrin's acridine ring are susceptible to nucleophilic substitution as revealed by the studies with dithiothreitol. Additionally, although most products from the reduction of ledakrin were extremely unstable, 1-aminoacridinone, produced enzymatically and with dithiothreitol, exhibited persistent stability under the studied conditions.

Comparison of aromatic and tertiary amine N-oxides of acridine DNA intercalators as bioreductive drugs. Cytotoxicity, DNA binding, cellular uptake, and metabolism.

Some N-oxide derivatives of DNA intercalators are bioreductive prodrugs that are selectively toxic under hypoxic conditions. The hypoxic selectivity is considered to result from an increase in DNA binding affinity when the N-oxide moiety is reduced. This study investigated whether differences in DNA binding affinity between N-oxides and their corresponding amines, measured by equilibrium dialysis, can account for the hypoxic cytotoxicity ratios (HCR) of tertiary amine N-oxide (-tO) and aromatic N-oxide (-aO) derivatives of the 1-nitroacridine nitracrine (NC) and its non-nitro analogue 9-[3-(N,N-dimethylamino)propylamino]acridine (DAPA). Cytotoxicity was measured in aerobic and hypoxic suspensions of Chinese hamster ovary (CHO) AA8 cells by clonogenic assay. HCR were much greater for NC-tO (820-fold) than for NC (5-fold) or NC-aO (4-fold), whereas DAPA and its N-oxides lacked hypoxic selectivity (1-fold). DNA binding measurements demonstrated that binding affinity is lowered more by aromatic than tertiary amine (side-chain) N-oxides, an observation that does not correlate with HCR. Compounds were accumulated in cells to high concentrations (C(i)/C(e) approximately 10-200), with the exception of the tertiary amine N-oxides, for which the ratio of intracellular to extracellular drug was less than unity. For NC-tO this probably resulted from low pK(a) values for both the acridine chromophore and the side-chain, whereas DAPA-tO may be too hydrophilic for efficient membrane permeation. Bioreductive drug metabolism, assessed by HPLC, was faster for the NC than the DAPA N-oxides. The high HCR of NC-tO relative to NC-aO is ascribed to the rapid and selective reduction of its N-oxide moiety, followed by activation of the NC intermediate by O(2)-sensitive reduction of its 1-nitro group to the corresponding 1-amine. The metabolism studies suggest that unmasking of DNA binding affinity by reductive removal of the N-oxide moiety, although not the only determinant, is important and needs to occur before nitroreduction for optimal effect.

Transcriptional template activity of covalently modified DNA.

The transcriptional template activity of covalent modified DNA is compared. 8-Methoxypsoralen (MOP), 3,4'dimethyl-8-methoxypsoralen (DMMOP) and benzopsoralen (BP) forming with DNA covalent complexes upon UV irradiation and exhibiting preference to pyrimidines, mostly thymines, differ in their cross-linking potency. MOP and DMMOP form both monoadducts and diadducts while no cross-links are formed by BP. Nitracrine (NC) forms covalent complexes with DNA upon reductive activation with dithiothreitol exhibiting a preference to purines and low cross-linking potency. Semilogarithmic plots of the relative template activity against the number of the drugs molecules covalently bound per 10(3) DNA nucleotides fit to regression lines corresponding to one-hit inactivation characteristics. The number of drug molecules decreasing RNA synthesis to 37% differ from 0.25 to 1.26 depending on the template used and the base preference but no dependence on the cross-linking potency was found.

32P-post-labelling analysis of nucleobases involved in the formation of DNA adducts by antitumor 1-nitroacridines.

Adducts generated in vitro by the reaction of 1-nitroacridines with poly(dN)s in the presence of dithiothreitol were used to identify a kind of nucleic base involved in the formation of individual adducts. The patterns of chromatographic spots corresponding to modified nucleotides obtained by 32P-post-labelling assay for synthetic homopolymers of four deoxyribonucleotides were compared with the fingerprints detected in the case of calf thymus DNA reacted with 1-nitroacridines under conditions in which the formation of identical DNA adducts as in cellular models was demonstrated in earlier investigations. Both compounds studied (Ledakrin and C-857) turned out to bind covalently only with purine nucleotides. Ledakrin formed with dG four and C-857 five different adducts. All of them were also detected in ctDNA. The incubation with poly(dA) resulted in four Ledakrin-dA species, two of which were found in ctDNA, and in two C-857-dA adducts that were not, however, observed in DNA containing samples. Modification of purines accounted for all adducts observed in ctDNA. For both compounds studied, the level of total binding to poly(dA) was about one order of magnitude lower than to poly(dG) for which it was comparable with the extent of ctDNA modification. This indicates that dG represents a preferential site of covalent binding of 1-nitroacridines to DNA.

In vitro DNA crosslinking by Ledakrin, an antitumor derivative of 1-nitro-9-aminoacridine.

Using agarose gel electrophoresis we confirmed that Ledakrin is capable of incurring covalent crosslinking in pBR322 plasmid DNA and also in poly(dGdC) in the presence of a simple activating system containing DTT. The identification of adducts resulting from DNA crosslinking was carried out by 32P-post-labelling assay. We assumed that such adduct(s) should be brought about more readily with double-stranded than with single-stranded polynucleotides or nucleotides. Since our earlier experiments had shown that guanine is a major site of covalent binding of 1-nitroacridines, we compared DNA adduct formation by Ledakrin for ctDNA, dG-containing synthetic homopolymers and 3'-pdG. 32P-Post-labelling assay revealed two adduct spots that were enhanced in samples containing double-stranded substrates in which interstrand crosslinking between guanines was possible, namely ctDNA and poly(dGdC).

Hypoxia-selective antitumor agents. 13. Effects of acridine substitution on the hypoxia-selective cytotoxicity and metabolic reduction of the bis-bioreductive agent nitracrine N-oxide.

A series of nuclear-substituted derivatives of nitracrine N-oxide (2; a bis-bioreductive hypoxia-selective cytotoxin) were prepared and evaluated, seeking analogues of lower nitroacridine reduction potential. Disubstitution with Me or OMe groups at the 4- and 5-positions did not provide analogues with one-electron reduction potentials significantly lower than those of the corresponding monosubstituted derivatives (E(1) ca. -350 mV for both the 4-OMe and 4,5-diOMe compounds). This appears not to be due to a concomitant raising of the acridine pKa but to a lack of direct electronic effect of substituents in the ring not bearing the nitro group. Conversely, placing two OMe groups in the nitro-bearing ring does result in a substantial further lowering of reduction potential (the 2,4-diOMe analogue has an E(1) of -401 mV). The mono- and disubstituted N-oxides have substantially lower cytotoxicities than the parent nitracrine N-oxide 2 but generally retain very high hypoxic selectivity. The OMe-substituted N-oxides all showed greater metabolic stability than 2 in hypoxic AA8 cell cultures, and the 4-OMe compound 6 had improved activity in EMT6 multicellular spheroids suggesting that this metabolic stabilization may allow more efficient diffusion in tumor tissue. The parent compound 2 was selectively toxic to hypoxic cells in KHT tumors in vivo and clearly superior to nitracrine itself (although only at doses which would eventually be lethal to the host). The analogues of lower E(1), including 6, were not superior to 2 in vivo, indicating that metabolic stabilization of the nitro group is not alone sufficient to improve therapeutic utility.

Effect of substituents on the metabolism of nitracrine in rat.

1. Male Wistar rats were treated with either the antitumour agent nitracrine (1-nitro-9-(3'-dimethylamino-N-propylamino)-acridine; NC), 4-methoxy-NC, NC-aliphatic-N-oxide, 4-methoxy-NC-aliphatic-N-oxide, or NC-aromatic-N-oxide (30 mumol/kg, via the femoral vein) and the major biliary and urinary metabolites analysed by hplc. 2. No NC or 4-methoxy-NC were detected in bile or urine of rat treated with NC or 4-methoxy-NC respectively, whereas the aliphatic N-oxides of NC and 4-methoxy-NC were recovered largely unchanged in both bile and urine. 3. NC-aromatic-N-oxide was rapidly and extensively converted to a major polar biliary product. This product was also synthesised enzymatically from NC-aromatic-N-oxide using rat liver cytosol and has been identified by mass and 1H-nmr spectrometry as 1-(S-glutathionyl)-9-(3'-dimethylamino-N-propylamino)-acridine-N(10)-oxi de. 4. The equivalent 1-(S-glutathionyl) conjugate appears to be formed from NC, and excreted in bile as a minor product, but not from 4-methoxy-NC. Further experiments with cytosol indicate that direct displacement of the nitro group by GSH is mediated by GSH transferase. 5. Finally, the major biliary metabolite of NC has been provisionally identified as a glucuronide of 1-nitro-2-hydroxy-NC. 6. It is concluded that, for at least a significant fraction of NC, nitroreduction does not occur. Further, N-oxidation of the aliphatic (but not the aromatic ring) nitrogen, plus 4-methoxy substitution, decreases the overall metabolism of NC in the rat.

An ESR and spectrophotometric study of the denitration of nitroheterocyclic drugs by liver homogenates and their metabolic consumption by liver microsomes from cytochrome P-450-induced mice.

This work extends a previous study on the mechanism of hepatic denitration of two nitroheterocyclic drugs (NHCD), quinifuryl and nitracrine, in which the release of nitric oxide (NO) from these compounds can be accompanied by the formation of a NO-heme iron complex. Pretreating mice with three inducers of cytochrome P-450 (phenobarbital, clophen A50 and butylated hydroxytoluene (BHT)) increased the yield of the nitrosyl complex which correlated with a rise in the cytochrome P-450 content of mouse liver microsomes. In contrast, treating the animals with beta-naphthoflavone decreased the complex yield while still increasing P-450 content. Treating the animals with any of the above inducers significantly increased the rate of NHCD metabolism in mouse liver microsomes. Based on these results, a possible mechanism for hepatic NHCD denitration is discussed.

5-Nitro-4-(N,N-dimethylaminopropylamino)quinoline (5-nitraquine), a new DNA-affinic hypoxic cell radiosensitizer and bioreductive agent: comparison with nitracrine.

Targeting of electron-affinic radiosensitizers to DNA via noncovalent binding (e.g., intercalation) may offer the potential for increasing sensitizing efficiency. However, it has been suggested that high-affinity DNA binding may compromise sensitization by restricting the mobility of sensitizers along the DNA, and by decreasing rates of extravascular diffusion in tumors. The weak DNA intercalator nitracrine (1-NC) is a more efficient radiosensitizer than related nitroacridines with higher DNA-binding affinities (Roberts et al., Radiat. Res. 123, 153-164, 1990). The present study investigates whether electron-affinic agents of even lower DNA-binding affinity may be superior to nitroacridines. The quinoline analog of 1-NC, 5-nitraquine (5-NO), was shown to have an intrinsic association constant for calf thymus DNA in 20 mM phosphate buffer which was 12-fold lower than that of 1-NC. 5-Nitraquine was not accumulated as efficiently as 1-NC by AA8 cells, but, despite a similar one-electron reduction potential, was 2- to 3-fold more potent than 1-NC as a hypoxia-selective radiosensitizer in vitro when compared on the basis of average intracellular concentration. Thus the radiosensitizing potency of 5-NQ appears not to be compromised by its low DNA-binding affinity. The cytotoxic mechanisms of 5-NQ and 1-NC appear to be similar (hypoxia-selective formation of DNA monoadducts), but 5-NQ is 1200-fold less potent than 1-NC as a cytotoxin. Despite this advantage, 5-NQ was not active in vivo as a radiosensitizer in SCCVII tumors. This lack of activity appears to be due to its relatively high toxicity in vivo (intraperitoneal LD50 of 105 mumol kg-1 in C3H/HeN mice), high one-electron reduction potential (-286 mV), and rapid metabolism to the corresponding amine in mice. The in vitro therapeutic index (hypoxic radiosensitizing potency/aerobic cytotoxic potency) of this weak DNA binder was lower than that of the non-DNA targeted radiosensitizer misonidazole, suggesting that DNA targeting enhances cytotoxicity more than radiosensitization. Development of useful DNA-targeted radiosensitizers may require the exploitation of DNA binding modes different from those of the nitroacridines and nitroquinolines.

Correlation of the radiosensitization potency afforded by nitroacridine intercalators with their electron scavenging efficiency in DNA.

Nitracrine (1-NC) and its nitro-positional analogues are electron-affinic DNA intercalating compounds that act as hypoxic cell radiosensitizers in cell culture, but are too rapidly metabolized to provide radiosensitization in vivo. We have explored the electron-trapping efficiencies of nitroacridines to see if such compounds can scavenge radicals formed in irradiated DNA and if the efficiency of such trapping is related to their radiosensitization properties. We have shown that a correlation does indeed exist as 1-NC, the most potent radiosensitizer, scavenges approximately 45% of the migrating electrons (formed by attachment of eaq- to the DNA) compared with approximately 4% for 4-NC, the least efficient radiosensitizer. The quenching of the delayed fluorescence from acridine orange bound to DNA was also used to probe intracellular uptake and association with DNA. The results are in accord with earlier measurements of average uptake and a suggestion that intercalators which form DNA complexes with short residence times will be preferable to highly bound agents as radiosensitizers. Since 1-NC possesses the smallest association constant with DNA of the four compounds, it is suggested that its high radiosensitization potency may well be related to it being able to interact with more radical sites on the DNA than the other analogues, in addition to its capture of migrating electrons in DNA.

32P-post-labeling analysis of DNA adduct formation by antitumor drug nitracrine (Ledakrin) and other nitroacridines in different biological systems.

A 32P-post-labeling method has been employed to detect DNA adducts formed by derivatives of nitro-9-aminoacridine in both cellular and non-cellular systems. The treatment of HeLa S3 cells in culture or Ehrlich ascites tumor cells in vivo with nitracrine and two other antitumor 1-nitro-9-aminoacridines, denoted C-857 and C-1006, resulted in covalent binding of these compounds to cellular DNA. Each derivative studied gave rise to a distinct pattern of adduct spots and the similarity of the respective adduct profiles was noted for the both cellular models. Calf thymus DNA samples modified in vitro with nitracrine and C-857 in the presence of either rat hepatic microsomal fraction or dithiothreitol yielded chromatographic profiles resembling those obtained in the cellular systems, suggesting similarity in the DNA adduct structures. There were also neither qualitative nor quantitative differences in calf thymus DNA modification by these two 1-nitro derivatives between aerobic and anaerobic conditions, thus the reduction of a nitro group seems not to be the only determinant of covalent binding to DNA in vitro. No DNA adduct formation was detected in the cellular systems used with 2-nitro and 4-nitro isomers of nitracrine that are devoid of cytotoxic activity, which provides further evidence that both covalent binding and DNA crosslinking, but not intercalation, are responsible for cytotoxic and antitumor properties of 1-nitro-9-aminoacridines.

In vitro binding of nitracrine to DNA in chromatin.

In the presence of sulfhydryl compounds nitracrine, an anticancer drug, binds covalently to DNA. The accessibility of DNA in chromatin both to nitracrine and to 8-methoxypsoralen, which was used as a reference compound in this study, when assayed in NaCl concentrations from 0 to 2 M show similar characteristics. The initial decrease reaches a minimum at 0.15 M NaCl above which dissociation of non-histone proteins and histones at higher ionic strengths is demonstrated by an increase in accessible sites. The relative accessibility of DNA in chromatin to nitracrine is, however, lower than that found for 8-methoxypsoralen. Partial dissociation of chromatin with 0.7 M NaCl increases the accessibility of DNA in chromatin when assayed in the absence of NaCl but has no apparent influence when estimated at ionic strength close to physiological conditions.

Hypoxia-selective antitumor agents. 1. Relationships between structure, redox properties and hypoxia-selective cytotoxicity for 4-substituted derivatives of nitracrine.

The nitroacridine derivative 9-[[3-(dimethylamino)propyl]amino]-1-nitroacridine (nitracrine) is selectively cytotoxic to hypoxic tumor cells in culture. However, the compound undergoes reductive metabolism too rapidly, with the reduction not being sufficiently inhibited by molecular oxygen in aerobic tissues, for it to demonstrate the same activity in vivo. In a search for derivatives with lower reduction potentials, we have synthesized and evaluated a series of derivatives bearing 4-substituents with a wide range of electronic properties. The one-electron reduction potentials (E(1] of these compounds, when compared under conditions of equivalent ionization, were highly correlated with sigma p values. However, at pH 7 the influence of substituent electronic properties was modified by prototrophic equilibria, with the basic nature of the acridine limiting the extent to which ring substituent electronic effects can be used to modulate reduction potential of the 1-nitro group. Nevertheless, comparison of the kinetics of the killing of AA8 cells under hypoxia suggests that some metabolic stabilization of the compounds can be achieved by the use of electron-donating substituents, with such compounds retaining the hypoxia-selective toxicity of nitracrine in cell culture. However, the 4-substituted nitracrines show no clear relationship between E(1) and cytotoxic potency, in distinct contrast to simpler nitroheterocycles such as nitroimidazoles.

Hypoxia-selective antitumor agents. 2. Electronic effects of 4-substituents on the mechanisms of cytotoxicity and metabolic stability of nitracrine derivatives.

The mechanism of cytotoxicity of a series of 4-substituted derivatives of 9-[[3-(dimethylamino)propyl]amino]-1-nitroacridine (nitracrine) has been studied, using a panel of DNA repair-defective mutants of the Chinese hamster ovary cell line AA8. Cell lines UV-4 and UV-5 were hypersensitive to nitracrine, with sensitivities approximately 10-fold greater than that of AA8, while EM-9 showed a hypersensitivity factor (HF) of about 2-fold. This pattern suggests the major cytotoxic lesions induced by nitracrine are bulky DNA monoadducts, rather than DNA interstrand cross-links as previously suggested. The desnitro analogue of nitracrine, which retains the intercalative potential of the latter but cannot be metabolically activated by nitro reduction, showed no hypersensitivity, indicating the specificity with which this panel of cell lines can discriminate different types of DNA damage. Several of the highly cytotoxic 4-substituted nitracrine derivatives showed HFs similar to that of the parent, but the less potent 4-dialkylamino and 4-COOMe derivatives showed much lower HFs for UV-4, suggesting that different mechanisms of cytotoxicity contribute. All compounds showed similar HFs under both aerobic and hypoxic conditions, indicating that hypoxia-selective toxicity in this series is due to a quantitative rather than qualitative change in the presence of oxygen. Rates of metabolic consumption of the compounds were measured under both aerobic and hypoxic conditions by bioassay against the sensitive UV-4 cell line. The results agreed well with previous inferences on metabolic stability derived from cell-killing kinetics and showed that electron-donating 4-substituents can be used to increase metabolic stability in vitro. Such stabilization may enhance the therapeutic utility of the nitroacridines in cancer therapy since rapid metabolism of nitracrine appears to prevent its activity against hypoxic cells in solid tumors.

Hypoxia-selective radiosensitization of mammalian cells by nitracrine, an electron-affinic DNA intercalator.

The radiosensitizing ability of the 1-nitroacridine nitracrine (NC) is of interest since it is an example of a DNA intercalating agent with an electron-affinic nitro group. NC radiosensitization was evaluated in Chinese hamster ovary cell (AA8) cultures at 4 degrees C in order to suppress the rapid metabolism and potent cytotoxicity of the drug. Under hypoxic conditions, submicromolar concentrations of NC resulted in sensitization (SER = 1.6 at 1 mumol dm-3). Sensitization was also seen under aerobic conditions but a concentration more than 10-fold higher was required. In aerobic cultures NC radiosensitization was independent of whether cells were exposed before and during, or after, irradiation. Postirradiation sensitization was not observed under hypoxic conditions. The time dependence of NC uptake and the development of radiosensitization were similar (maximal at 30 min at 4 degrees C under hypoxia) suggesting that sensitization, unlike cytotoxicity, is due to unmetabolized drug. NC is about 1700 times more potent as a radiosensitizer than misonidazole. This high potency is adequately accounted for by the electron affinity of NC (E(1) value at pH7 of -275 mV versus NHE) and by its accumulation in cells to give intracellular concentrations approximately 30 times greater than in the medium. However, concentrations of free NC appear to be low in AA8 cells, presumably because of DNA binding. If radiosensitization by NC is due to bound rather than free drug, it suggests that intercalated NC can interact very efficiently with DNA target radicals. This is despite a binding ratio in the cell estimated as less than 1 NC molecule/400 base pairs under conditions providing efficient sensitization. This work suggests a new approach in the search for more effective clinical radiosensitizers, and poses questions on the means by which intercalated drugs can interact with DNA damage.

Mutagenicity of nitracrine analogues in Salmonella typhimurium: mutational specificity and activation by bacterial enzymes and rat-liver S9.

The mutagenic potential of 9-[(3-dimethylaminopropyl)amino]-acridine and its 1-(nitracrine), 2-, 3- and 4-nitro derivatives was studied in several strains of Salmonella typhimurium, using a plate-incorporation assay. In view of the potential importance of DNA binding and nitro group reduction, binding constants and redox potentials were determined. The parent compound was mutagenic only in the frameshift tester strain TA1537, and this effect was not enhanced by the plasmid pKM101. Each of the nitroarenes showed significant activity in S. typhimurium strains TA1537 and TA1538, and this mutagenicity was enhanced by the presence of plasmid pKM101. All caused reversion of the base-pair substitution allele in hisG46 in strain TA100 and this effect was either largely or totally dependent upon the plasmid. The frameshift mutagenic effects of the 3-nitro and 4-nitro compounds appeared to be little dependent upon the classical nitroreductase which is deficient in strain TA98NR, or transacetylase enzyme lacking in strain TA98/1,8-DNP6, whereas the activity of the 1-nitro compound depended partly on that nitroreductase enzyme, and that of the 2-nitro compound on having both activities present. In the latter two cases, the mutagenic effects of the compounds could not be restored by the addition of an S9 mix of mammalian enzymes. Mutagenicity data were compared with physicochemical parameters. The results do not distinguish between the view that the orientation of the nitro group with respect to the aromatic plane dictates mutagenic potential, and the earlier view that the nitro group redox potential is important. However, the extraordinary mutagenic potency of the 1-NO2 derivative is not explainable by its physicochemical properties alone.