Bis-benzimidazole anticancer agents: targeting human tumour helicases.

Certain DNA minor groove binding agents, distamycin, netropsin, and a series of anticancer bis-benzimidazoles can block DNA helicase activity by binding to duplex DNA at specific base sequences. DNA helicases are crucial to cell DNA replication, transcription and repair because these enzymes separate double-stranded DNA, thereby preparing the strands for enzymatic manipulation. From our studies we have developed a hypothesis that focuses on cellular DNA helicase action as a mechanistic site where these minor groove binders can act. A crucial aspect for modulation of DNA activity by drugs is for specificity and selectivity. A series of DNA-interactive bis-benzimidazole analogues of Hoechst 33258 was also prepared to explore the potential for anticancer activity mediated for certain of the drugs via bioreductive activation by endogenous NADH or NADPH. The biological endpoints examined included intracellular distribution in euoxic and hypoxic conditions observed by fluorescence microscopy; relative efficacy as antimetabolites determined by the MTT [tetrazolium salt, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide] assay in euoxic and hypoxic conditions; and relative inhibitory activities on human DNA helicase, as determined by degree of dissociation of GC B6486 DNA. The intracellular distribution was unique to each of the test compounds. Compounds V-93 and V-153, the respective semiquinone and quinone derivatives, demonstrated the predicted enhanced cytotoxicity and anti-helicase activities, supporting the concept that preferential binding of DNA at 5'-CG and TG sequences provides a novel approach to anticancer drug development.

Antihelicase action of CI-958, a new drug for prostate cancer.

CI-958, a new DNA-intercalating drug derived from a series of substituted 2H-[1] benzothiopyrano[4,3,2-cd]indazoles, is being tested in clinical trails because of its curative properties against murine solid tumor models and because it has demonstrated activity in a pilot phase II study of patients with hormone-refractory prostate cancer. However, the mechanism of anticancer action of CI-958 has not been established. Because CI-958 binds to DNA and DNA helicases are profoundly affected by DNA-binding drugs, we examined the effects of CI-958 on human DNA helicase action. DNA helicase activity was measured by strand dissociation of double-stranded (ds) DNA with a gel electrophoresis assay, and ATPase activities were determined on thin-layer chromatography by measurement of the conversion of ATP to ADP. For human helicase blockade, CI-958 is slightly more potent than doxorubicin (EC50 values 0.17 and 0.26 microM, respectively). We observed no difference in helicase-blockade EC50 values recorded for three helicase substrates containing A-T rich, G-C rich, and both types of oligonucleotide sequences. The effects of CI-958 helicase blockade and DNA-dependent ATPase activities were similar for the two reactions. The kinetics of the blockade by CI-958 of the human DNA helicase indicates that it involves a reversible ternary complex of helicase-drug-dsDNA. CI-958 produces potent blockade of human DNA helicases with no apparent strong DNA sequence-binding preference. Similar potency against helicase strand dissociation and DNA-dependent ATPase suggests that the mechanism against these reactions is the same. The blockade of DNA helicases by CI-958 may be central in its mechanism of action as an anticancer drug.

Anthracycline antibiotic blockade of SV40 T antigen helicase action.

We previously showed that anthracycline antibiotics potently block SV40 large T antigen helicase; in the present study, we describe the kinetics and the structure-activity characteristics of this process. The concentration vs effect data for helicase blockade were fitted by the Hill equation to yield nearly parallel log-concentration effect curves for a series of active anthracycline antibiotics. The effective concentration for 50% helicase blockade (EC50) values ranged from 0.34 microM for daunorubicin to 40.8 microM for 3'-deaminodaunorubicin. Clinically inactive 3'-N-acyl anthracyclines produced no blockade. The Hill constants for the blockade ranged from 1.1 to 1.6 for the entire series of active anthracyclines, indicating no positive cooperativity and suggesting that a single molecule of bound drug is sufficient to block helicase action. The EC50 values for several clinically effective anthracyclines showed a relationship to the average DNA binding constants for these drugs, and Lineweaver-Burk analysis of the blockade kinetics indicated non-competitive inhibition. The kinetics of the blockade indicated that the anthracycline, DNA, and helicase form a ternary complex that is irreversible under the reaction conditions. This mechanism may be central to the cytotoxic and anti-cancer activities of anthracycline antibiotics and may be useful in understanding the enzymatic mechanism of DNA helicase action.

Antihelicase action of DNA-binding anticancer agents: relationship to guanosine-cytidine intercalator binding.

DNA-binding antibiotics such as intercalators, narrow groove binders, and other substances modify duplex DNA, making it an altered substrate for DNA helicases. The intercalators daunorubicin, actinomycin D, echinomycin, and elsamicin, the narrow groove binders distamycin and mithramycin, and the plant toxin teniposide, each representing a different chemical class, block SV40 large T antigen DNA helicase action with IC50 values ranging from 4 x 10(-8) to 2 x 10(-6) M. A partially purified human HeLa cell DNA helicase is also potently blocked by daunorubicin, distamycin, and teniposide. Because eukaryotic cells contain helicases of varying abundance, specificity, and type, this site of action for DNA-binding antibiotics may help explain antibiotic potency and specificity for DNA or RNA inhibition. The antihelicase effect of the antibiotic-double-stranded DNA complex may be central to the anticancer activities of these substances. An additional interesting correlation is the antihelicase action of DNA-intercalating antibiotics and their DNA-binding preference for G-C base pair sites. The G-C base pair binding preference of the intercalating antibiotics may result from evolutionary selection because of the higher G-C binding stability, compared with A-T binding stability. The combination of the higher base pair stability at G-C regions and increased duplex DNA stability induced by intercalating antibiotic yields a total additive stability of the intercalator-G-C base pair complex that resists helicase action.

Helicase inhibition by anthracycline anticancer agents.

Helicases are essential to both DNA replication and transcription because they separate double-stranded DNA, preparing the single strands for replication or transcription. Because the anti-cancer anthracycline antibiotics stabilize double-stranded DNA primarily by their intercalative binding, we expected the intercalated antibiotics to interfere with helicase action. We examined anthracycline antibiotic effects on SV40 large T antigen helicase activity, using a duplex DNA helicase substrate of 32P-labeled 17-mer annealed to complementary M13mp19(+) circular single-stranded DNA. The T antigen helicase activity was potently inhibited by the anthracycline antibiotics. The T antigen helicase IC50 values for the anthracycline antibiotics were as follows: nogalamycin, 2 x 10(-7) M; daunorubicin, 4 x 10(-7) M; doxorubicin, 4 x 10(-7) M; idarubicin, 1.8 x 10(-6) M; 4'-epidoxorubicin, 2 x 10(-6) M; aclacinomycin, 4 x 10(-6) M; and menogaril, 6 x 10(-6) M. Partially purified helicases from HeLa cells and murine mammary carcinoma FM3A cells also were potently inhibited by doxorubicin, with IC50 values of 4 x 10(-7) M and 9 x 10(-7) M, respectively. Because the abundance, specificities, and types of helicases vary in the cell, this site of action for anthracycline antibiotics may help explain anthracycline potency, drug specificity for DNA or RNA inhibition, and some types of cellular resistance to these drugs.

Phase I clinical and pharmacokinetic study of zeniplatin, a new platinum complex.

Forty-six patients with refractory solid malignancies received the new platinum complex [2,2-bis(aminomethyl)-1,3-propanediol-N-N'] [1,1-cyclobutanedicarboxylato] [(2-)0,0')] platinum (zeniplatin). Zeniplatin was given, without hydration or mannitol, as a 60- to 90-min i.v. infusion every 3 weeks at doses ranging from 8 to 145 mg/m2. The maximum tolerated dose of zeniplatin was 145 mg/m2. The dose-limiting toxicity of zeniplatin was dose-related leukopenia and neutropenia, with the nadir usually observed between 1 and 2 weeks after therapy and recovery usually occurring by 3 weeks after therapy. Thrombocytopenia was rare. The most prominent non-hematological side-effect of zeniplatin was nausea and vomiting. Other non-hematological side-effects were mild or absent. Zeniplatin did not induce significant neurological or auditory toxicity. Zeniplatin was not nephrotoxic at doses less than or equal to 120 mg/m2. At 145 mg/m2, the clearance decreased by a mean of 40% after 2 cycles of therapy. Two patients, one with malignant melanoma and one with renal cell cancer, achieved a partial response. Pharmacokinetics of free (plasma ultrafiltrates) and total platinum in plasma were determined in 5 patients. An in vitro study of the rate and extent of zeniplatin binding to protein in human plasma was also performed. Free and total platinum were measured by flameless atomic absorption spectrometry; free zeniplatin was measured in ultrafiltrate by HPLC. Total and free plasma platinum concentrations were co-modelled using the information from the in vitro study.(ABSTRACT TRUNCATED AT 250 WORDS)

Induction of a human carbonyl reductase gene located on chromosome 21.

Carbonyl reductase (EC 1.1.1.184) belongs to the group of enzymes called aldo-keto reductases. It is a NADPH-dependent cytosolic protein with specificity for many carbonyl compounds including the antitumor anthracycline antibiotics, daunorubicin and doxorubicin. Human carbonyl reductase was cloned from a breast cancer cell line (MCF-7). The cDNA clone contained 1219 base paires with an open reading frame corresponding to 277 amino acids encoding a protein of Mr 30,375. Southern analysis of genomic DNA digested with several restriction enzymes and analyzed by hybridization with a labeled cDNA probe indicated that carbonyl reductase is probably coded by a single gene and does not belong to a family of structurally similar enzymes. Southern analysis of 17 mouse/human somatic cell hybrids showed that carbonyl reductase is located on chromosome 21. Carbonyl reductase mRNA could be induced 3-4-fold in 24 h with 10 microM 2,(3)-t-butyl-4-hydroxyanisole (BHA), beta-naphthoflavone or Sudan 1.

Experimental chemotherapy-induced skin necrosis in swine. Mechanistic studies of anthracycline antibiotic toxicity and protection with a radical dimer compound.

The reactivity of antitumor anthracycline and mitomycin C antibiotics with the oxomorpholinyl radical dimers, bi(3,5,5-trimethyl-2-oxomorpholin-3-yl) (TM3) and bi(3,5-dimethyl-5-hydroxymethyl-2-oxomorpholin-3-yl) (DHM3), was studied in vitro. The oxomorpholinyl radical reduced daunorubicin to a quinone methide intermediate that reacted with solvent to form 7-deoxydaunorubicinone. The solvolysis reaction followed first order kinetics, and the reactivity rate constants (k2) measured for seven anthracycline analogues ranged from 2 X 10(-2) s-1 to 8.0 X 10(-4) s-1. The chemical reactivity of each anthracycline quinone methide correlated with the total skin toxicity caused by the respective parent anthracycline following injection into swine skin. Microscopic examination of experimental lesions in swine skin resemble those observed in humans after inadvertant chemotherapy extravasation. Hydrocortisone sodium succinate was not effective for the treatment of doxorubicin-induced skin necrosis, whereas DHM3 was effective for the treatment of skin necrosis caused by all seven anthracyclines and by the quinone containing antibiotic, mitomycin C.

Loss of fluorescence by anthracycline antibiotics: effects of xanthine oxidase and identification of the nonfluorescent metabolites.

Rat liver cytosol and buttermilk xanthine oxidase both converted 7-deoxypyrromycinone, the 7-deoxyaglycone of marcellomycin, a new anthracycline antibiotic, to a nonfluorescent compound under anaerobic conditions and in the presence of an electron donor. Reduced nicotinamide adenine dinucleotide and reduced nicotinamide adenine dinucleotide phosphate were equally effective electron donors for liver cytosol, and xanthine was the best cofactor for xanthine oxidase. However, xanthine was inactive with liver cytosol. Reactions with xanthine oxidase obeyed Menten-Michaelis kinetics and were inhibited by allopurinol. No xanthine oxidase activity was detected in liver cytosol. Xanthine oxidase also induced a loss of fluorescence when incubated with 7-deoxydaunorubicin aglycone. The nonfluorescent metabolite of 7-deoxypyrromycinone was tentatively identified as the dihydroquinonic derivative of the parent deoxyaglycone on the basis of its spectrophotometric, fluorescent, thin layer chromatographic, and mass spectral characteristics. Our data demonstrate that more than one enzymatic activity, xanthine oxidase, and an unidentified rat liver cytosolic enzyme convert the 7-deoxyaglycones of anthracycline antibiotics to nonfluorescent metabolites.

Comparative anthracycline metabolism in rats: loss of marcellomycin fluorescence.

The in vitro metabolism of marcellomycin by rat tissue fractions showed conversion of marcellomycin to 7-deoxypyrromycinone, bisanhydropyrromycinone, and an as yet unidentified compound by rat liver homogenate, microsomes, cytosol, and mitochondria, and purified hepatic reduced nicotinamide adenine dinucleotide phosphate-cytochrome P-450 reductase, under anaerobic conditions and in the presence of reduced nicotinamide adenine dinucleotide phosphate. All these fractions except the purified reductase subsequently induced a progressive loss of fluorescence. Mitochondria, however, were much less active than microsomes, cytosol, and homogenate in inducing this latter phenomenon. Marcellomycin was converted to 7-deoxyaglycones only partially by nuclei. No loss of fluorescence was observed with this subcellular fraction. No loss of fluorescence was observed when doxorubicin or daunorubicin were incubated under similar conditions. The appearance of a compound with distinct spectrophotometric properties was demonstrated by absorbance spectrometry. The formation of a compound with different fluorescent characteristics was excluded, as was the binding of the aglycones to subcellular components. The activity inducing the loss of fluorescence was studied in greater detail with cytosol. It predominated in the liver and required both an electron donor and anaerobic conditions. The optimal pH for the reaction was between 7.5 and 8.0. Our results suggest the existence of an enzymatic pathway capable of converting the fluorescent nucleus of marcellomycin to a nonfluorescent metabolite.

DNA alkylation by enzyme-activated mitomycin C.

After anaerobic reductive activation by either NADPH cytochrome P-450 reductase (EC 1.6.2.4) or xanthine oxidase (EC 1.2.3.2), mitomycin C readily alkylated DNA. When the mitomycin C-alkylated DNA is digested by DNase, snake venom phosphodiasterase, and alkaline phosphatase, only partial release of the monofunctionally linked mitomycin C nucleotide adduct occurs. Cross-linked adducts are not released into dinucleotides but resist nuclease digestion and remain in oligonucleotides and insoluble precipitates. Kinetic analyses show that the nuclease-resistant fraction which is indicative of DNA cross-linking by mitomycin C takes place quite readily. This nuclease-resistant fraction is particularly significant when the amount of total bound mitomycin C is less than 15 mumol/mmol of DNA. The cross-linked mitomycin C product accounts for more than half of the total alkylation under all pH conditions tested. Our data suggest that particular DNA sites are available for DNA cross-linking by mitomycin C, and these sites are probably the preferred and immediate alkylating targets. Furthermore, DNA cross-links by mitomycin C are not the secondary product of monofunctional adducts. Activity of both flavoenzymes is pH dependent, hence, mitomycin C activation and the rate of DNA alkylation are pH dependent. At elevated mitomycin C alkylation of DNA, the highest amount of cross-linking occurs at neutral pH. High pressure liquid chromatographic separation of the nuclease-digested DNA detected one major and two less prominent mitomycin C adducts. These were verified to be mononucleotide mitosene types by UV spectra showing maximum absorbance at 312 and 250 nm. The major adduct was purified and identified as O6-(2'-deoxyguanosyl)-2,7-diaminomitosene by NMR, indicating that the O6 position of guanine is a preferred site in DNA for at least monofunctional linkage formation.

Effect of hyperthermia on the in vitro metabolism of doxorubicin.

The effect of hyperthermia on the uptake and metabolism of doxorubicin (ADM) was studied in in vitro systems. ADM uptake in rat liver slices was not affected by increasing the temperature from 37 degrees C to 43 degrees C. In rat liver homogenates, the aerobic transformation of ADM was low and was not affected by hyperthermia. Approximately 90% of the parent drug remained unchanged after 60 minutes of incubation, and two metabolites, adriamycinol and a polar metabolite, were formed in small amounts. Under anaerobic conditions, ADM was quickly and extensively converted to two metabolites identified as 7-deoxyadriamycinol aglycone and 7-deoxyadriamycin aglycone. Whereas the disappearance of ADM and the formation of 7-deoxyadriamycin aglycone were not modified by the hyperthermic conditions, there was a slight but significant increase of the formation of 7-deoxyadriamycinol alycone (area under the concentration versus time curve in microM X minute: 216 +/- 24 at 37 degrees C; 235 +/- 24 at 39.5 degrees C; and 271 +/- 8 at 42 degrees C; P less than or equal to 0.05). However, the percentage of dA3 was not significantly different at the end of the incubation.

Interactions between cyclophosphamide and doxorubicin metabolism in rats. II. Effect of cyclophosphamide on the aldoketoreductase system.

Under anaerobic conditions, in comparison to liver microsomes obtained from normal controls, liver microsomes obtained from rats pretreated with cyclophosphamide formed significantly less 7-deoxydoxorubicinol aglycone (P less than or equal to .05), whereas the disappearance of doxorubicin and the formation of 7-deoxydoxorubicin aglycone were unaffected. When directly investigated, the reduction of 7-deoxydoxorubicin aglycone to 7-deoxydoxorubicinol aglycone by microsomes was inhibited by cyclophosphamide pretreatment. Liver cytosols from controls and cyclophosphamide-treated rats reduced daunorubicin to daunorubicinol and 7-deoxydoxorubicin aglycone to 7-deoxydoxorubicinol aglycone at the same rate, which indicates the lack of effect of cyclophosphamide pretreatment on the cytosolic aldoketoreductase. The results suggest the existence of a microsomal carbonyl reduction system for anthracycline antibiotics and indicate that cyclophosphamide does affect the metabolism of doxorubicin; in rats, this interaction results only in an alteration of the relative concentrations of presumably inactive metabolites, the 7-deoxyaglycones. The importance of these findings for the pharmacological interaction between doxorubicin and cyclophosphamide in humans remains to be investigated.

Cellular pharmacology and antitumor activity of N-(p-azidobenzoyl)-daunorubicin, a photoactive anthracycline analogue.

We have previously utilized N-(p-azidobenzoyl)daunorubicin (NABD), a photoactive analogue of daunorubicin (DNR), to identify unique anthracycline-binding polypeptides in rodent tissues and in tumor cells. Using cultured P388 tumor cells, we have now compared the cellular pharmacology and antitumor activity of NABD with that of DNR. Although rapidly accumulated by cells, the intracellular concentration of NABD was less than 20% that of DNR at steady-state levels. The cellular uptake of both drugs by P388 cells was dependent on extracellular drug concentration in the medium and on temperature. The rapid efflux of NABD and DNR from P388 cells in drug-free medium was reduced at lowered temperature (0 degrees C). Cytofluorescence microscopy demonstrated that NABD was predominantly localized in the cytoplasm, in contrast to the nuclear localization of DNR. NABD produced dose-dependent inhibition of [3H]thymidine (IC50 = 10.0 microM) and [3H]uridine (IC50 = 1.60 microM) incorporation in P388 cells to a lesser degree than DNR ([3H]thymidine, IC50 = 0.15 microM and [3H]uridine, IC50 = 0.70 microM). Continuous exposure to NABD inhibited P388 cell proliferation with an IC50 of 0.27 microM, compared with an IC50 of 0.017 microM for DNR. NABD is a pharmacologically active, photoactive analogue of DNR, which possesses properties different from those of the parent drug but similar to those of other anthracycline analogues. Photoaffinity labeling studies with NABD may identify important cytoplasmic constituents which interact with this type of anthracycline and perhaps with the anthracycline antibiotics in general.

Doxorubicin-induced skin necrosis in the swine model: protection with a novel radical dimer.

The treatment of doxorubicin (DOX) extravasation tissue injury is poorly defined. A swine model has been developed to study DOX skin toxicity and potential pharmacologic antidotes. Intradermal injections of DOX in miniature female weanling swine produced predictable and dose-dependent ulcerations that closely resemble lesions observed in humans following extravasation of DOX. The ulcers reached maximal size at 3 weeks following DOX administration and were completely healed by 7 weeks. Bi(3,5-dimethyl-5-hydroxymethyl-2-oxomorpholin-3-yl) (DHM3) is a radical dimer that can react with DOX in vitro to produce deoxydoxorubicin aglycone, an inactive anthracycline metabolite. When DHM3 was administered into the same intradermal injection site 15 minutes after DOX, the maximum ulcer size was reduced 80%, and the healing time was reduced to 5 weeks. The protection from toxicity was highly dependent on the time interval between DOX and DHM3 injections, with no protection noted after a 60-minute interval. Our data verify the swine model as a useful tool to study DOX-induced extravasation injury. Furthermore, DHM3 is an effective antidote for DOX-induced skin necrosis and has potential for clinical use.

Radical dimer rescue of toxicity and improved therapeutic index of adriamycin in tumor-bearing mice.

The product of adriamycin (ADR) reductive glycosidic cleavage is the pharmacologically inactive 7-deoxyadriamycin aglycone. Bi(3,5-dimethyl-5-hydroxymethyl-2-oxomorpholin-3-yl) (DHM3) is a radical dimer which reacts with ADR in vitro to produce this aglycone. We utilized DHM3 to prevent ADR toxicity in mice. CD2F1 male mice were given a single dose of ADR, 25 mg/kg i.p., which was acutely lethal as indicated by a median survival time of 7 days. DHM3 administered as a single i.p. dose of 50 mg/kg 15 or 30 min following ADR provided significant protection with median survival times greater than 9 wk. Mice bearing ascitic L1210 leukemic cells were given ADR, 0, 6.6, 15, or 25 mg/kg i.p. 1 day following inoculation of tumor. DHM3 administered as a single 50 mg/kg i.p. dose 20 min after ADR had no significant effect on ADR efficacy at the lower dose range (% treated versus control = 171 and 285 for 6.6 and 15.0 mg/kg, respectively). Less than 15% of the animals in these treatment groups were long-term survivors. However, following high doses of ADR (25 mg/kg), DHM3 protected mice from ADR lethality and over 70% of animals were long-term survivors. The determination of parent ADR and ADR aglycone content in several tissues indicated that the concentration of ADR was reduced in those animals that received DHM3 15 min after ADR. Correspondingly an increase in ADR aglycone concentration in each tissue resulted from DHM3 treatment. DHM3 represents a novel class of compounds that can ameliorate ADR toxicity and has potential use as a rescue agent.

Doxorubicin and cisplatin excretion into human milk.

Plasma and milk concentrations of doxorubicin (DOX) and cisplatin were measured after iv administration of these agents to a lactating patient with ovarian cancer. Cisplatin was undetectable in human milk. Milk concentrations of DOX often exceeded those detected in concomitant plasma samples. For DOX, the highest milk:plasma concentration ratio was 4.43:1 and was observed 24 hours after administration of the drug. The area under concentration versus time curve (AUC) for DOX was approximately the same in plasma and milk. Doxorubicinol was the major metabolite of DOX in plasma and in milk. The AUC for doxorubicinol was ten times higher in milk than in plasma. However, the total amount of anthracycline antibiotic delivered in the milk (maximum concentration of active anthracycline antibiotic: 0.24 mg/L) was negligible.

Metabolism and disposition of menogaril (NSC 269148) in the rabbit.

We have investigated the metabolism and disposition, in rabbits, of menogaril (7-OMEN), a new anthracycline antibiotic recently introduced into clinical trials. 7-OMEN was administered by rapid i.v. injection at a dosage of 2.5 mg/kg. 7-OMEN and metabolites were assayed by high performance liquid chromatography. Plasma concentrations of 7-OMEN declined in biexponential fashion with a terminal half-life of 2.7 h. The area under the plasma concentration versus time curve was 1.3 microM X h. The systemic clearance of 7-OMEN was 57.6 ml/min/kg. No metabolite of 7-OMEN was detected in plasma. At 8 h after treatment, the cumulative urinary and biliary excretions of 7-OMEN equivalents amounted to 1.3 and 3.4% of the total administered dose, respectively. 7-OMEN was the predominant fluorescent compound in urine, but four metabolites were also seen. In bile, 7-OMEN represented only 9.6% of the cumulative excretion and six metabolites were observed. Among the organs, lungs contained the highest concentrations of parent drug. Substantial concentrations of metabolites were observed in the kidneys, liver, duodenum, and small intestine. Three of the observed metabolites of 7-OMEN have been tentatively identified as N-demethylmenogaril, 7-deoxynogarol, and N-demethyl-7-deoxynogarol.