DNA sequence-specific adducts of adriamycin and mitomycin C

Inhibition of Autophagy Promotes the Elimination of Liver Cancer Stem Cells by CD133 Aptamer-Targeted Delivery of Doxorubicin

Doxorubicin is the most frequently used chemotherapeutic agent for the treatment of hepatocellular carcinoma. However, one major obstacle to the effective management of liver cancer is the drug resistance derived from the cancer stem cells. Herein, we employed a CD133 aptamer for targeted delivery of doxorubicin into liver cancer stem cells to overcome chemoresistance. Furthermore, we explored the efficacy of autophagy inhibition to sensitize liver cancer stem cells to the treatment of CD133 aptamer-doxorubicin conjugates based on the previous observation that doxorubicin contributes to the survival of liver cancer stem cells by activating autophagy. The kinetics and thermodynamics of aptamer-doxorubicin binding, autophagy induction, cell apoptosis, and self-renewal of liver cancer stem cells were studied using isothermal titration calorimetry, Western blot analysis, annexin V assay, and tumorsphere formation assay. The aptamer-cell binding andintracellular accumulation of doxorubicin were quantified via flow cytometry. CD133 aptamer-guided delivery of doxorubicin resulted in a higher doxorubicin concentration in the liver cancer stem cells. The combinatorial treatment strategy of CD133 aptamer-doxorubicin conjugates and an autophagy inhibitor led to an over 10-fold higher elimination of liver cancer stem cells than that of free doxorubicin in vitro. Future exploration of cancer stem cell-targeted delivery of doxorubicin in conjunction with autophagy inhibition in vivo may well lead to improved outcomes in the treatment of hepatocellular carcinoma.

Design, synthesis, and DNA sequence selectivity of formaldehyde-mediated DNA-adducts of the novel N-(4-aminobutyl) acridine-4-carboxamide

A novel derivative of the anti-tumor agent N-[2-(dimethylamino)ethyl]acridine-4-carboxamide (DACA) was prepared by reduction of 9-oxoacridan-4-carboxylic acid to acridine-4-carboxylic acid with subsequent conversion to N-(4-aminobutyl)acridine-4-carboxamide (C4-DACA). Molecular modeling studies suggested that a DACA analogue comprising a side chain length of four carbons was optimal to form formaldehyde-mediated drug-DNA adducts via the minor groove. An in vitro transcription assay revealed that formaldehyde-mediated C4-DACA-DNA adducts selectively formed at CpG and CpA dinucleotide sequences, which is strikingly similar to that of formaldehyde-activated anthracenediones such as pixantrone.

Site-specific DNA-doxorubicin conjugates display enhanced cytotoxicity to breast cancer cells

Doxorubicin (Dox) is widely used for breast cancer treatment but causes serious side effects including cardiotoxicity that may adversely impact patient lifespan even if treatment is successful. Herein, we describe selective conjugation of Dox to a single site in a DNA hairpin resulting in a highly stable complex that enables Dox to be used more effectively. Selective conjugation of Dox to G15 in the hairpin loop was verified using site-specific labeling with [2-¹⁵N]-2′-deoxyguanosine in conjunction with [¹H–¹⁵N] 2D NMR, while 1:1 stoichiometry for the conjugate was validated by ESI-QTOF mass spectrometry and UV spectroscopy. Molecular modeling indicated covalently bound Dox also intercalated into the stem of the hairpin and stability studies demonstrated the resulting Dox-conjugated hairpin (DCH) complex had a half-life >30 h, considerably longer than alternative covalent and noncovalent complexes. Secondary conjugation of DCH with folic acid (FA) resulted in increased internalization into breast cancer cells. The dual conjugate, DCH-FA, can be used for safer and more effective chemotherapy with Dox and this conjugation strategy can be expanded to include additional anticancer drugs.

Genotoxicity of doxorubicin in F344 rats by combining the comet assay, flow-cytometric peripheral blood micronucleus test, and pathway-focused gene expression profiling

Doxorubicin (DOX) is an antineoplastic drug effective against many human malignancies. DOX's clinical efficacy is greatly limited because of severe cardiotoxicity. To evaluate if DOX is genotoxic in the heart, ~7-week-old, male F344 rats were administered intravenously 1, 2, and 3 mg/kg bw DOX at 0, 24, 48, and 69 hr and the Comet assays in heart, liver, kidney, and testis and micronucleus (MN) assay in the peripheral blood (PB) erythrocytes using flow cytometry were conducted. Rats were euthanized at 72 hr and PB was removed for the MN assay and single cells were isolated from multiple tissues for the Comet assays. None of the doses of DOX induced a significant DNA damage in any of the tissues examined by the alkaline Comet assay. Contrastingly, the glycosylase enzymes-modified Comet assay showed a significant dose dependent increase in the oxidative DNA damage in the cardiac tissue (P ≤ 0.05). In the liver, only the top dose induced significant increase in the oxidative DNA damage (P ≤ 0.05). The histopathology showed no severe cardiotoxicity but non-neoplastic lesions were present in both untreated and treated samples. A severe toxicity likely occurred in the bone marrow because no viable reticulocytes could be screened for the MN assay. Gene expression profiling of the heart tissues showed a significant alteration in the expression of 11 DNA damage and repair genes. These results suggest that DOX is genotoxic in the heart and the DNA damage may be induced primarily via the production of reactive oxygen species.

Stability and iron coordination in DNA adducts of Anthracycline based anti-cancer drugs

There is evidence that the interaction of the α-ketol group of the Doxorubicin and Epirubicin anti-cancer drugs with Fe(III) generates hydroxyl radicals under aerobic conditions, causing cardiotoxicity in patients. Considering that the formation of DNA adducts is one of the main targets of Anthracycline drugs, we have in the present study characterized several [Anthracycline-DNA]Fe(III) complexes with respect to their stability and Fe(III) coordination, by means of MD simulations. Iron is found to coordinate well to the drugs containing an α-ketol group, this being the only group of the drug that binds to the metal. The complexes containing an α-ketol group, [Doxorubicin-DNA]Fe(III) and [Epirubicin-DNA]Fe(III), thus show greater stability than those not containing it, i.e., [Daunorubicin-DNA]Fe(III), [Idarubicin-DNA]Fe(III) and [5-Imino-Daunorubicin]Fe(III). Metal attachment to the α-ketol group is furthermore facilitated by the phosphate groups of DNA. The coordination to iron in the [Doxorubicin-DNA]Fe(III) system is smaller than that found for the [Epirubicin-DNA]Fe(III) system, and the corresponding number of coordinating waters in the former is larger than in the latter. This may in turn result in higher hydroxyl radical production, thus explaining the increased cardiotoxicity noted for Doxorubicin.

Reactions of glyceraldehyde 3-phosphate dehydrogenase sulfhydryl groups with bis-electrophiles produce DNA-protein cross-links but not mutations

The environmental contaminant 1,2-dibromoethane and diepoxybutane, an oxidation product of the important industrial chemical butadiene, are bis-functional electrophiles and are known to be mutagenic and carcinogenic. One mechanism by which bis-electrophiles can exert their toxic effects is through the induction of genotoxic and mutagenic DNA-peptide cross-links. This mechanism has been shown in systems overexpressing the DNA repair protein O6 -alkylguanine DNA-alkyltransferase (AGT) or glutathione S-transferase and involves reactions with nucleophilic cysteine residues. The hypothesis that DNA-protein cross-link formation is a more general mechanism for genotoxicity by bis-electrophiles was investigated by screening nuclear proteins for reactivity with model monofunctional electrophiles. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was identified as a candidate because of the nucleophilicity of two cysteine residues (Cys152 and Cys246) in reaction screens with model electrophiles (Dennehy, M. K. et al. (2006) Chem. Res. Toxicol. 19, 20-29). Incubation of GAPDH with bis-electrophiles resulted in inhibition of its catalytic activity, but only at high concentrations of diepoxybutane. In vitro assays indicated DNA-GAPDH cross-link formation in the presence of diepoxybutane, and bis-electrophile reactivity at Cys246 was confirmed using mass spectral analysis. In contrast to AGT, overexpression of human GAPDH in Escherichia coli did not enhance mutagenesis by diepoxybutane. We propose that the lack of mutational enhancement is in part due to the inherently lower reactivity of GAPDH toward bis-electrophiles as well as the reduced DNA binding ability relative to AGT, preventing the in vivo formation of DNA-protein cross-links and enhanced mutagenesis.

An evaluation of the role of insulin-like growth factors (IGF) and of type-I IGF receptor signalling in hepatocarcinogenesis and in the resistance of hepatocarcinoma cells against drug-induced apoptosis

Strong evidence emphasizes the role of the insulin-like growth factor (IGF) system and of type-I IGF receptor (IGF-IR) signalling in tumourigenesis. In this connection: (i) changes in the expression pattern of components of the IGF system (autocrine/paracrine expression of IGF-I and -II, overexpression of IGF-IR, decreased expression of IGF-binding proteins (IGFBPs) and of type-II IGF receptor/cation-independent mannose-6-phosphate receptor (IGF-II/M6PR) and (ii) increased serum concentrations of proteases that cleave the IGFBPs (e.g., cathepsin D) were observed in patients with hepatocellular carcinomas (HCC), in human hepatoma cell lines and in their conditioned culture medium, as well as in rodent models of hepatocarcinogenesis. Accordingly, studies carried out with animal models do suggest that the IGF system and IGF-IR signalling may play a role in hepatocarcinogenesis and in deregulated proliferation and apoptosis of HCC cells. Finally the instrumental role of Raf/MEK/ERK, one of the signalling cascades stimulated by IGF-IR, in anthracycline-induced apoptosis of HepG2 and Huh-7 human hepatoma cell lines emphasizes that care must be taken when designing combinations of antitumoural molecules for antineoplastic treatment. This review addresses the putative roles of the IGF system in primary HCC, with a special focus on the underlying molecular mechanisms. In a second part it emphasizes the putative interference of IGF-IR signalling with chemotherapeutic drug-induced apoptosis in human hepatoma cells.

Tumor suppression and therapy sensitization of localized and metastatic breast cancer by adenovirus p53

We have examined the effects of a replication-defective adenovirus encoding p53 (RPR/INGN 201 [Ad5CMV-p53]; Adp53), alone or in combination with the breast cancer therapeutic doxorubicin (Adriamycin), to suppress growth and induce apoptosis in breast cancer cells in vitro. We have also examined the in vivo effect of intratumoral administration of Adp53, alone or in combination with doxorubicin, to suppress the growth of established subcutaneous MDA-MB-435 breast cancer tumors. Finally, using the MDA-MB-435 orthotopic model of metastatic breast cancer, we have examined the effect of systemic administration of Adp53, alone or in combination with doxorubicin, to reduce the incidence of metastases. We find that whereas in vitro treatment of cells with Adp53 reduces [(3)H]thymidine incorporation by about 90% at 48 hr, cell viability at 6 days is reduced by only some 50% relative to controls. Although apoptosis is detectable in Adp53-treated cultures, these results suggest that a large fraction of Adp53-treated cells merely undergo reversible cell cycle arrest. Combined treatment with Adp53 and doxorubicin results in a greater than additive loss of viability in vitro and increased apoptosis. In vivo, locally administered Adp53 suppresses growth of established subcutaneous tumors in nude mice and suppression is enhanced by doxorubicin. In the metastatic breast cancer model, systemic administration of Adp53 plus doxorubicin leads to a significant reduction in the incidence of metastases relative to Adp53 or doxorubicin alone. Taken together, these data indicate an additive to synergistic effect of Adp53 and doxorubicin for the treatment of primary and metastatic breast cancer.

Interstrand cross-linking by adriamycin in nuclear and mitochondrial DNA of MCF-7 cells

Activation of Adriamycin by formaldehyde leads to the formation of drug-DNA adducts in vitro and these adducts stabilise the DNA to such a degree that they function as virtual interstrand cross-links. The formation of these virtual interstrand cross-links by Adriamycin was investigated in MCF-7 cells using a gene-specific interstrand cross-linking assay. Cross-linking was measured in both the nuclear-encoded DHFR gene and in mitochondrial DNA (mtDNA). Cross-link formation increased linearly with Adriamycin concentration following a 4 h exposure to the drug. The rate of formation of Adriamycin cross-links in each of the genomes was similar, reaching maximal levels of 0.55 and 0.4 cross-links/10 kb in the DHFR gene and mtDNA respectively, following exposure to 20 micro M Adriamycin for 8 h. The interstrand cross-link was short lived in both DNA compartments, with a half-life of 4.5 and 3.3 h in the DHFR gene and mtDNA respectively. The kinetics of total Adriamycin adduct formation, detected using [(14)C]Adriamycin, was similar to that of cross-link formation. Maximal adduct levels (30 lesions/10 kb) were observed following incubation at 20 micro M drug for 8 h. The formation of such high levels of adducts and cross-links could therefore be expected to contribute to the mechanism of action of Adriamycin.

A survey of the sequence-specific interaction of damaging agents with DNA: emphasis on antitumor agents

This article reviews the literature concerning the sequence specificity of DNA-damaging agents. DNA-damaging agents are widely used in cancer chemotherapy. It is important to understand fully the determinants of DNA sequence specificity so that more effective DNA-damaging agents can be developed as antitumor drugs. There are five main methods of DNA sequence specificity analysis: cleavage of end-labeled fragments, linear amplification with Taq DNA polymerase, ligation-mediated polymerase chain reaction (PCR), single-strand ligation PCR, and footprinting. The DNA sequence specificity in purified DNA and in intact mammalian cells is reviewed for several classes of DNA-damaging agent. These include agents that form covalent adducts with DNA, free radical generators, topoisomerase inhibitors, intercalators and minor groove binders, enzymes, and electromagnetic radiation. The main sites of adduct formation are at the N-7 of guanine in the major groove of DNA and the N-3 of adenine in the minor groove, whereas free radical generators abstract hydrogen from the deoxyribose sugar and topoisomerase inhibitors cause enzyme-DNA cross-links to form. Several issues involved in the determination of the DNA sequence specificity are discussed. The future directions of the field, with respect to cancer chemotherapy, are also examined.

Daunomycin-induced unfolding and aggregation of chromatin

Using equilibrium dialysis and sedimentation velocity analysis, we have characterized the binding of the anti-tumor drug daunomycin to chicken erythrocyte chromatin before and after depletion of linker histones and to its constitutive DNA under several ionic strengths (5, 25, and 75 mM NaCl). The equilibrium dialysis experiments reveal that the drug binds cooperatively to both the chromatin fractions and to the DNA counterpart within the range of ionic strength used in this study. A significant decrease in the binding affinity was observed at 75 mM NaCl. At any given salt concentration, daunomycin exhibits higher binding affinity for DNA than for linker histone-depleted chromatin or chromatin (in decreasing order). Binding of daunomycin to DNA does not significantly affect the sedimentation coefficient of the molecule. This is in contrast to binding to chromatin and to its linker histone-depleted counterpart. In these instances, preferential binding of the drug to the linker DNA regions induces an unfolding of the chromatin fiber that is followed by aggregation, presumably because of histone-DNA interfiber interactions.

A critical evaluation of the mechanisms of action proposed for the antitumor effects of the anthracycline antibiotics adriamycin and daunorubicin

The mechanisms responsible for the antiproliferative and cytotoxic effects of the anthracycline antibiotics doxorubicin (Adriamycin) and daunorubicin (daunomycin) have been the subject of considerable controversy. This commentary addresses the potential role of DNA synthesis inhibition, free radical formation and lipid peroxidation, DNA binding and alkylation, DNA cross-linking, interference with DNA strand separation and helicase activity, direct membrane effects, and the initiation of DNA damage via the inhibition of topoisomerase II in the interaction of these drugs with the tumor cell. One premise underlying this analysis is that only studies utilizing drug concentrations that reflect the plasma levels in the patient after either bolus administration or continuous infusion are considered to reflect the basis for drug action in the clinic. The role of free radicals in anthracycline cardiotoxicity is also discussed.

Observation of daunomycin and nogalamycin complexes with duplex DNA using electrospray ionisation mass spectrometry

The noncovalent binding of the antitumour drugs daunomycin and nogalamycin to duplex DNA has been studied using electrospray ionisation mass spectrometry (ESI-MS). The conditions for the preparation of drug/duplex DNA complexes and for their detection by ESI-MS have been optimised. Ions corresponding to these complexes were most abundant relative to free DNA when prepared in the pH range 8-9, and using gentle ESI interface conditions. Self-complementary oligonucleotides, 5'-d(GGCTAGCC)-3' or 5'-d(CGGCGCCG)-3', annealed in the presence of a 5-fold molar excess of either nogalamycin or daunomycin gave ESI mass spectra in which the most intense ions corresponded to three molecules of drug bound to duplex DNA, with some evidence for four drug molecules bound. For binding to 5'-d(TGAGCTAGCTCA)(2)-3', complexes containing up to four nogalamycin and six daunomycin molecules were observed. These data are consistent with the neighbour exclusion principle whereby intercalation occurs between every other base pair such that up to four bound drugs would be expected for the 8 mers and up to six for the 12 mer. Competition experiments involving a single drug in an equimolar mixture of two oligonucleotides (5'-d(TGAGCTAGCTCA)(2)-3' with either 5'-d(CGGCGCCG)(2)-3' or 5'-d(GGCTAGCC)(2)-3') showed ions arising from complexes of drug/5'-d(CGGCGCCG)(2)-3' were more intense than complexes of drug/5'-d(GGCTAGCC)(2)-3', relative to those from the 12 mer in each mixture. While this suggests ESI-MS has the potential to detect differences in sequence selectivity, more detailed experiments involving a comparison of the relative ionisation efficiency of different oligonucleotides and a wider range of intercalators are required to establish this definitively. ESI mass spectra from experiments in which both drugs were reacted with the same oligonucleotide were more complex, such that a clear preference for one drug could not be established.

Characterization of covalent adriamycin-DNA adducts

Adriamycin is a popular antineoplastic agent whose ability to form covalent adducts with DNA has been correlated to cellular apoptosis (programmed cell death) in tumor models. We have isolated and purified this adduct formed under oxido-reductive (Fenton) conditions in Tris buffer. We show by homo- and heteronuclear NMR spectroscopy that the covalent Adriamycin-DNA adduct is structurally equivalent to that resulting from direct reaction with formaldehyde. Covalent linkage of the drug to one of the DNA strands confers remarkable stability to the duplex, indicated by a 162-fold reduction in the rate of strand displacement compared with the complex with noncovalently bound drug. Glyceraldehyde also engenders covalent Adriamycin-DNA complexes, providing a possible relevant biological context for in vivo adduct formation.

Enzymatic activation of DNA cleavage by dynemicin A and synthetic analogs

Dynemicin A (1), a member of the enediyne family of natural products, binds to double-stranded DNA (K(B) approximately 10(4) M(-1)) and in the presence of millimolar concentrations of a reducing cofactor such as NADPH or GSH reacts to cleave DNA. In this work, we show that the two flavin-based enzymes ferredoxin-NADP+ reductase and xanthine oxidase catalyze the reductive activation of 1 by NADPH and NADH, respectively. The enzyme-catalyzed reductive activation of 1 leads to more rapid and efficient cleavage of DNA, even with 10-20-fold lower concentrations of the stoichiometric reductant. Significantly, the enzymatic systems are also found to activate the tight-binding (K(B) > or = 10(6) M(-1)) synthetic dynemicin analogs 3 and 5 toward DNA cleavage. These same analogs do not undergo reductive activation with NADPH or NADH alone, where evidence has been obtained to support the proposal that the DNA-bound drugs are protected from reductive activation. The new enzymatic activation processes described may have important implications for chemistry occurring with 1 and synthetic analogs in vivo, as well as for the future development of dynemicin-based anticancer agents.

Inhibition of bacteriophage T7 RNA polymerase in vitro transcription by DNA-binding pyrrolo[2,1-c][1,4]benzodiazepines

The interactions of several pyrrolo[2, 1-c][1,4]benzodiazepine (PBD) antitumor antibiotics with linearized plasmid pGEM-2-N-ras DNA have been analyzed by quantitative in vitro transcription (QIVT) and in vitro transcription footprinting (IVTF) methods. A concentration-dependent inhibitory effect of the PBDs on transcription is observed using both techniques. The rank order for overall inhibition of transcription by the QIVT method is found to be: sibiromycin > tomaymycin > anthramycin > DC-81 > neothramycin, whereas the IVTF experiments show a different ranking: sibiromycin > anthramycin > neothramycin > tomaymycin. In addition, stimulation of transcription was observed at low PBD concentrations in both the QIVT and IVTF experiments. These results demonstrate unequivocally that the formation of PBD-DNA adducts at AGA-5' base sequences on the transcribed strand results in transcription blockage for all PBDs examined. Furthermore, the sequence of flanking base pairs appears to influence the degree of blocking, with the sequences ACAGAAA-5', AAAGATG-5', AGAGATA-5', and CAAGAAC-5' providing the most pronounced blocks for all PBDs studied in this system. Neothramycin and tomaymycin cause additional blocks at some GGA-5' and TGA-5' sequences. Parallel MPE-Fe(II) footprinting studies have revealed PBD binding sites on both the transcribing and nontranscribing strands, although all transcription blocks determined from the IVTF assays are due to drug bound on the transcribing DNA template strand.

Comparison of DNA sequence selectivity of anthracycline antibiotics and their 3'-hydroxylated analogs

The sequence selectivity of three anthracyclines and their 3' hydroxylated analogs (in which an OH replaces NH3+ in the daunosamine at neutral pH) was examined in DNase I footprinting experiments on a 158-bp DNA fragment. We found that chemical modification of the daunosamine at C3' has more drastic consequences for sequence selectivity than chemical modification at C4 and C14 of the aglycone moiety. All anthracyclines and hydroxylated derivatives selectively recognize the triplet PyAPy. The importance of NH3+ in stabilizing the interaction was evidenced. First of all, comparable protection patterns require 5 times more hydroxyanthracycline than regular anthracycline. Furthermore, it is only after the replacement of NH3+ by OH that an additional protection site - CGC--appears. GGC is the site of best selectivity of the hydroxyanthracyclines. Anthracyclines can be considered both intercalators (aglycone moiety) and minor groove binders (sugar moiety). Since intercalating drugs show a slight preference for GC base pairs, we suggest hydroxylated anthracyclines to have a sequence specificity closer that of pure intercalators. Chemical modifications at C4 and C14 only modify the hydrogen bonding stabilization of the DNA-aglycone moiety complex: the more the anthracycline or its analog is lipophilic, the less it will interact with the sugar-phosphate chain.

Epidermal growth factor, phorbol esters, and aurintricarboxylic acid are survival factors for MDA-231 cells exposed to adriamycin

The ability of epidermal growth factor (EGF), insulin-like growth factor-1 (IGF-1), insulin, 12-O-tetradecanoylphorbol-13-acetate (TPA), and aurintricarboxylic acid (ATA) to protect the human breast cancer cell line MDA-231 from death induced by the anticancer drug adriamycin was investigated. Cell death was induced in the MDA-231 cells either by a short-time exposure to a high dose of adriamycin (2 micrograms.ml-1.1h-1) and further culturing in the absence of the drug, or by continuous exposure to a low dose of adriamycin (0.3 micrograms/ml). Cell death was evaluated after 48 h of incubation by several techniques (trypan blue dye exclusion, lactic dehydrogenase activity, cellular ATP content, transmission electron microscopy, and DNA fragmentation). EGF, TPA, and ATA, each at an optimal concentration of 20 ng/ml, 5 ng/ml, and 100 micrograms/ml respectively, substantially enhanced survival of cells exposed either to a high or low dose of adriamycin. Neither IGF-1 nor insulin, each at concentrations of 20 ng/ml, had an effect on cell survival. The three survival factors enhanced protein synthesis in the untreated cells and attenuated the continuous decrease in protein synthesis in the adriamycin-treated cells. Moreover, the three survival factors protected the MDA-231 cells from death in the absence of protein synthesis (cycloheximide 30 micrograms/ml). These results suggest that EGF, TPA, and ATA promote survival of adriamycin pretreated cells by at least two mechanisms: enhancement of protein synthesis and by a protein synthesis independent process, probably a posttranslational modification effect.

Adriamycin induces large deletions as a major type of mutation in CHO cells

Adriamycin (ADR), a commonly used cancer chemotherapy antibiotic, exhibits a variety of genotoxicities. In this study, we have examined the mutagenicity of ADR at the hypoxanthine-guanine phosphoribosyltransferase gene (hprt) in Chinese hamster ovary (CHO) cells and the xanthine-guanine phosphoribosyltransferase locus (gpt) in a pSV2gpt-transformed CHO cell line, AS52. Although ADR induced a dose-dependent increase of mutant frequency at both loci, it was more mutagenic to the gpt gene than to the hprt locus. Multiplex PCR analysis revealed that 35% of the 103 independent ADR-induced HPRT-deficient mutants carried large deletions. Among these deletion mutants, 33% were total gene deletions, 22% affected multiple exons, and 42% involved a single exon, of which most (9/15) were exon 1. The majority (63%) of ADR-induced AS52 mutants had a total deletion of the gpt gene. These observations indicate that ADR induces large deletions as a major type of gene mutation in mammalian cells, suggesting the involvement of reactive oxygen species as one mutagenic pathway in the mutagenesis of ADR.

Formation of adriamycin--DNA adducts in vitro

Adriamycin is known to induce the formation of adducts with DNA when reacted under in vitro transcription conditions. The factors affecting the extent of adduct formation were examined in order to establish the critical components and optimal conditions required for the reaction, and to gain insight into the nature of the DNA-adduct complex. There was a strong dependence on reaction temperature (with a 40-fold increase of adducts at 40-50 degrees C compared to 10 degrees C), pH (maximum adducts at pH 7), but little dependence on the oxygen level. There was an absolute requirement for a reducing agent, with adducts detected with DTT, beta-mercaptoethanol and glutathione, maximal adducts were formed at high levels of DTT (5-10 mM). Adducts were also formed with a xanthine oxidase/NADH reducing system, with increasing amounts of adducts detected with increasing NADH; no adducts were detected in the absence of either the enzyme or NADH. Of fourteen derivatives studied, only four yielded a similar extent of adduct formation as adriamycin; there was no absolute requirement for a carbonyl at C13 or hydroxyl at C14. Adducts were also observed with ssDNA but required a longer reaction time compared to dsDNA. The sequence specificity of adduct formation with ssDNA was examined using a primer-extension assay; almost all adducts were associated with a guanine residue. Overall, the results are consistent with a two-step reaction mechanism involving reductive activation of adriamycin, with the activated species then reacting with the guanine residues of either dsDNA or ssDNA.

Mutations in a shuttle vector exposed to activated mitomycin C

The cytotoxicity of the potent antibiotic and antitumor agent mitomycin C (MMC) is due to its irreversible binding to DNA. Alkylating species generated by bioreductive activation of MMC are known to cause monoadducts and cross-links in DNA by specifically binding to guanine residues. To gain insight into how these lesions lead to base- and sequence-specific mutations, shuttle vector pSP189 was treated with MMC chemically reduced by treatment with sodium borohydride, replicated in human Ad293 cells, rescued in bacteria, and analyzed for mutations in the supF tRNA gene sequence. The MMC-induced mutations were predominantly base substitutions. Eighty-four percent of the base substitutions were transversions, with G:C-->T:A the major transversion. Single base deletions were the other major mutational event, and 77% of these were G:C deletions. Base positions 115, 123, and 163 were mutational hot spots based on the frequency of independent mutations. Identification of a single MMC adduct (presumed to be a modified G on the basis of its Rf value) and clustering of MMC-induced mutations at three GC-rich areas (nt 100-123, 152-163, and 168-176) suggested that the mutational spectrum we found was due to binding of MMC to guanine on either strand of the plasmid DNA.

Does adriamycin induce interstrand cross-links in DNA?

Under nonenzymatic conditions in vitro, Adriamycin appears to form interstrand cross-links with DNA over 1-2 days. This is the first report of such Adriamycin-induced interstrand cross-links in vitro. Cross-links were measured by a fluorescence based renaturation assay and also by gel electrophoresis. Both procedures revealed an increase of cross-linking with reaction time and with increasing Adriamycin concentration and a 5-6-fold enhancement in the presence of Fe3+ ions. The cross-link contains the Adriamycin chromophore, with a lambda max of 508 nm, intercalated at the GpC site of cross-linking. Maximal stoichiometry of the cross-link was one per 11-20 bp. The cross-link appears to involve adducts of the Adriamycin chromophore linked to the N2 of guanine, with no indication that N7 of guanine is involved. Given that the mode of action of Adriamycin still remains obscure, even after 20 years of clinical use, the possibility that interstrand DNA cross-links may be associated with the clinical mechanism of action of this drug should now be fully addressed.

Excision repair reduces doxorubicin-induced genotoxicity

LacI mutations induced by doxorubicin in a wild-type, uvr(A)BC repair-proficient E. coli strain were analyzed by DNA sequencing. These mutations were contrasted with mutations previously recovered from doxorubicin-treated uvrB- organisms in order to assess the role of excision repair in doxorubicin-induced genotoxicity. After a 30-min exposure of wild-type E. coli to 330 microM doxorubicin, survival was 34% and the overall lacI mutation frequency increased 1.8-fold to 340 x 10(-8). The distribution of doxorubicin-induced mutants among subclasses of mutation involving the i-d and lac operator regions differed significantly between repair-proficient and -deficient strains. Distributional differences appeared to result both from a decrease in deletions involving the lac operator and an increase in base substitutions involving the i-d region in repair proficient organisms. However, elements of the doxorubicin-induced mutation spectrum in uvrB- E. coli are still discernable in wild-type organisms. These elements include the remarkable shift of 3'-deletion endpoints to palindromic sequence within the lac operator and the recovery of multiple isolates of T:A-->A:T transversions at position 96 in doxorubicin-treated cultures. These observations suggest that components of the uvr(A)BC nucleotide excision repair system function through a general mechanism prior to fixation of mutations to reduce, but not completely eliminate, the genotoxic effects of doxorubicin.

Application of high-performance liquid chromatography for recognition of covalent nucleic acid modification with anticancer drugs

Covalent modification of DNA by antineoplastic agents represents a potent biochemical lesion which can play a major role in drug mechanism of action. The ability to measure levels of DNA covalent modifications in target cells in vivo may, therefore, be seen as the ultimate form of therapeutic drug monitoring. Additionally, elucidation of the structure of critical DNA adducts and definition of their role in tumour cell cytotoxicity will provide more selective targets for rational drug design of new cancer chemotherapeutic agents. High-performance liquid chromatography has contributed significantly to all these areas. In vivo levels of nucleic acid covalent modifications are in the range of 1 in 10(5)-10(8) nucleotides precluding the use of conventional high-performance liquid chromatographic detection methods. Several classes of natural product anticancer drugs have been shown to bond covalently to nucleic acids under optimal laboratory conditions. These have proved more accessible to high-performance liquid chromatographic analysis because of their lipophilicity and strong UV chromophores. However, the majority of experimental evidence to date suggests that with the exception of mitomycin C and morpholino-anthracyclines these compounds do not exert their primary mechanism of action through nucleic acid covalent modification. DNA adducts of alkylating and platinating agents are more difficult to detect by high-performance liquid chromatography and can be chemically unstable. These compounds interact with DNA on the basis of chemical kinetics. Thus, the principle sites of attachment tend to be with the most nucleophilic base (guanine) at its most reactive centre (N-7 position). Limited in vivo high-performance liquid chromatographic studies with all classes of anticancer drugs indicate a much more complex pattern of adductation than would have been anticipated from in vitro studies alone. Some of these differences are probably due to methodological artefacts but these studies stress the need for sensitive detection methods and reliable sample preparation (nucleic acid extraction and digestion techniques) when attempting to determine nucleic acid covalent modifications by anticancer drugs.

Adriamycin promotes neurite outgrowth in the "neurite-minus" N1A-103 mouse neuroblastoma cell line

Adriamycin, an anticancer agent acting on topoisomerase II, promotes the arrest of cell division and neurite extension in a "neurite-minus" murine neuroblastoma cell line, N1A-103. This morphological differentiation is accompanied by a blockade in the S phase of the cell cycle, modification of the amount of peripherin, and appearance of the beta 7-tubulin isoform. Yet, adriamycin-induced N1A-103 cells fail to express other neuronal markers, such as long-lasting Ca2+ channels, synaptophysin, and the shift in the proportion of the beta'1 tubulin isoform to the beta'2 isoform, whose appearance parallels the terminal differentiation of the wild type neuroblastoma cell line N1E-115. Hence, a comparison of the behavior of these two cell lines leads to the proposal that there are two programs of neuroblastoma differentiation: one where expression is triggered by the arrest of cell division and which is observed in adriamycin-induced N1A-103 variant cells, and the other, presumably occurring further downstream, which would involve further changes in morphogenesis and acquisition of new electrophysiological properties.

Determination of covalent binding to intact DNA, RNA, and oligonucleotides by intercalating anticancer drugs using high-performance liquid chromatography. Studies with doxorubicin and NADPH cytochrome P-450 reductase

An HPLC method is described which can determine covalent binding to intact nucleic acid by intercalating anticancer drugs and at the same time remove noncovalently bound intercalated drug. The method uses a column containing a nonporous 2-microns DEAE anion-exchange resin capable of chromatographing nucleic acids greater than 50,000 bases in size in under 1 h. After priming with 1 mg of DNA, the column behaves as an intercalator affinity column, strongly retaining the drug while allowing the nucleic acid to pass through normally. Retained drug is released with an injection of 0.1 M potassium hydroxide. Incubations were performed with the intercalator doxorubicin, which is also believed to bind covalently to DNA. When [14C]doxorubicin was mixed with DNA, at a concentration where all the drug would bind by intercalation, the column retained 82% of the total radioactivity, only 18% migrated with the nucleic acid. If the DNA was mildly denatured by treatment with 2 M sodium chloride at 50 degrees C for 45 min before chromatography, then 99.8% of total radioactivity was retained, only background counts migrated with the nucleic acid, as was the case with single-stranded DNA and RNA without any treatment. Purified NADPH cytochrome P-450 reductase was used to activate doxorubicin. DNA inhibited the metabolism of the drug by the enzyme, no covalent binding occurred with RNA, low levels occurred with single-stranded DNA (34 pmol/100 micrograms), and the highest levels were recorded with oligonucleotides (243 pmol/100 micrograms). The assay was sufficiently sensitive to measure covalent binding to DNA extracted from MCF-7 human breast cancer cells treated with 50 microM [14C]doxorubicin (18.6 pmol/100 micrograms). Thus, covalent binding to DNA, RNA, and oligonucleotides by intercalators can be measured quickly (20 min) without the need to either digest the nucleic acid or subject it to long sample preparation techniques.

Induction of stable transcriptional blockage sites by adriamycin: GpC specificity of apparent adriamycin-DNA adducts and dependence on iron(III) ions

Initiated transcription complexes were exposed to adriamycin for up to 48 h. Subsequent elongation of the transcription complex revealed the presence of a series of discrete long-lived blockage sites. The mole fraction of blocked transcripts increased linearly with reaction time, adriamycin concentration, and Fe(III) concentration. Optimal conditions for formation of the blocked transcript were 24-h reaction time, 10 microM adriamycin, and 75 microM Fe(III) ions. Nine high-intensity blocked transcripts were observed, and all correspond to transcription proceeding up to G of GpC sequences of the nontemplate strand. The presence of 75 microM Fe(III) ions enhanced the amount of transcriptional blockages by 12-15-fold. Two blocked transcripts decayed with a half-life of 0.32 and 1.9 h, and one of these exhibited 100% effective delayed termination 6 bp downstream of the original blockage site. All other blockages were unchanged after 3 h of elongation. Bidirectional transcription footprinting was used to define the physical size of the drug-induced blocking moiety as a maximum of 2 bp, and this was observed at all three GpC elements probed by RNA polymerase from both directions. The nature of the apparent covalent adducts has not yet been established but is probably a G-specific adduct deriving from a reduced form of the drug (quinone methide). Although the GpC specificity suggests an interstrand G-drug-G cross-link, these were not detected by heat denaturation and subsequent denaturing gel electrophoresis of the end-labeled promoter fragment.