Topoisomerase inhibitors can selectively interfere with different stages of simian virus 40 DNA replication.

I have found that antineoplastic drugs which are known to be inhibitors of mammalian DNA topoisomerases have pronounced and selective effects on simian virus 40 DNA replication. Ellipticine, 4'-(9-acridinylamino)methanesulfon-m-aniside, and Adriamycin blocked decatenation of newly replicated simian virus 40 daughter chromosomes in vivo. The arrested decatenation intermediates produced by these drugs contained single-strand DNA breaks. Ellipticine in particular produced these catenated dimers rapidly and efficiently. Removal of the drug resulted in rapid reversal of the block and completion of decatenation. The demonstration that these drugs interfere with decatenation suggests that they may exert their cytotoxic and antineoplastic effects by preventing the separation of newly replicated cellular chromosomes. Camptothecin rapidly breaks replication forks in growing Cairns structures. It is likely that the target of camptothecin is the "swivel" topoisomerase required for DNA replication and that it is located at or very near the replication fork in vivo. Evidence is presented that many of the broken Cairns structures are in fact half-completed sister chromatid exchanges. One pathway for the resolution of these structures is completion of the sister chromatid exchange to produce a circular head-to-tail dimer.

Swiveling and decatenation of replicating simian virus 40 genomes in vivo.

We have found that type II topoisomerase inhibitors have two effects on replicating simian virus 40 genomes in vivo: production of catenated dimers and slowed replication of the last 5% of the genome. This suggests that type II topoisomerase simultaneously decatenates and facilitates replication fork movement at this stage of DNA replication. On the basis of this observation, a detailed model is proposed for the roles of topoisomerases I and II in simian virus 40 DNA replication.

Mechanism of action of the antileukemic xanthone psorospermin: DNA strand breaks, abasic sites, and protein-DNA cross-links.

Psorospermin, a cytotoxic dihydrofuranoxanthone isolated from Psorospermum febrifugum, produced aberrant simian virus 40 DNA replication intermediates when added to lytically infected CV-1 monkey kidney cells. The aberrant viral intermediates showed dose-dependent DNA strand breaks and protein-DNA cross-links, as well as decreased electrophoretic mobility. Simian virus 40 DNA from psorospermin-treated cells was shown to contain numerous abasic (apyrimidinic/apurinic) sites. The density of abasic sites was a function of the psorospermin dose. We conclude that psorospermin causes extensive loss of DNA bases in vivo. Primary amine groups of cellular proteins are known to react with abasic sites to form covalent protein-DNA cross-links and DNA strand breaks. Cytochrome c cross-linked spontaneously to viral DNA prepared from psorospermin-treated cells but not to DNA from untreated cells. This suggests that the protein-DNA cross-links and many of the DNA strand breaks observed in vivo result from reactions between abasic sites and chromosomal proteins. It is likely that the protein-DNA cross-links and DNA strand breaks contribute to the cytotoxicity and antineoplastic activity of psorospermin.

Chloroquinoxaline sulfonamide (NSC 339004) is a topoisomerase IIalpha/beta poison.

Chloroquinoxaline sulfonamide (chlorosulfaquinoxaline, CQS, NSC 339004) is active against murine and human solid tumors. On the basis of its structural similarity to the topoisomerase IIbeta-specific drug XK469, CQS was tested and found to be both a topoisomerase-IIalpha and a topoisomerase-IIbeta poison. Topoisomerase II poisoning by CQS is essentially undetectable in assays using the common protein denaturant SDS, but easily detectable with strong chaotropic protein denaturants. The finding that detection of topoisomerase poisoning can be so dependent on the protein denaturant used in the assay has implications for drug discovery efforts and for our understanding of topoisomerase poisons.

Unbalanced growth in mouse cells with amplified dhfr genes.

When grown in the absence of methotrexate, cells carrying unstably amplified dihydrofolate reductase (dhfr) genes have a growth disadvantage that is a function of their level of gene amplification. Although this growth disadvantage is thought to drive the loss of unstably amplified dhfr genes in the absence of methotrexate, its mechanism is not understood. The present studies of murine cell lines with different levels of dhfr gene amplification demonstrate that such cells experience increased unbalanced growth (excess RNA and protein content relative to DNA content) with increased levels of dhfr gene amplification. Stathmokinetic analysis of a cell line with unstably amplified dhfr genes showed that the unbalanced growth was associated with a very low rate of G1/S transit, which suggests that amplified DNA sequences may activate a cell cycle checkpoint at the G1/S boundary. Hydroxyurea, which is known to induce rapid elimination of amplified genes at sub-cytotoxic concentrations, also inhibits the cell cycle at the G1/S transition and causes unbalanced growth. Earlier work has shown that hydroxyurea selectively targets those cells within the heterogeneous drug resistant cell populations which have the highest amplified gene dosage. The finding that unstable gene amplification and hydroxyurea have similar effects on the cell suggests that hydroxyurea may achieve this selective targeting by pushing those cells with the highest levels of gene amplification over a critical stress threshold to cause growth arrest or cell death.

Inhibition of topoisomerase II by ICRF-193, the meso isomer of 2,3-bis(2,6-dioxopiperazin-4-yl)butane. Critical dependence on 2,3-butanediyl linker absolute configuration.

The bis(2,6-dioxopiperazine)s are a structurally and mechanistically unique class of topoisomerase II inhibitors that do not bind DNA and that do not stabilize topoisomerase II-DNA strand passing intermediates ("cleavable complexes"). The most effective topoisomerase II inhibitor in the bis(2,6-dioxopiperazine) series is ICRF-193 (meso or S*, R* isomer), with a meso 2,3-butanediyl linker connecting the dioxopiperazine rings. The two enantiomeric diastereomers, (R,R) and (S,S), of ICRF-193 possessing the two optically active 2,3-butanediyl linkers have been prepared from their respective optically pure 2,4-diaminobutanes via 2,3-diaminobutane-N,N,N',N'-tetraacetic acid, esterification, and imide formation. Both in vivo and in vitro assays for catalytic inhibition of topoisomerase II were employed to show that the (S,S)- and (R,R)-isomers are almost inactive as topoisomerase II inhibitors. The data indicate that the meso stereochemistry of the alkanediyl linker is crucial for activity and provides additional evidence that the cytotoxicity of the bis(2,6-dioxopiperazine)s is due to their ability to inhibit topoisomerase II.

Inhibition of topoisomerase II by liriodenine.

The cytotoxic oxoaporphine alkaloid liriodenine, isolated from Cananga odorata, was found to be a potent inhibitor of topoisomerase II (EC 5.99.1.3) both in vivo and in vitro. Liriodenine treatment of SV40 (simian virus 40)-infected CV-1 cells caused highly catenated SV40 daughter chromosomes, a signature of topoisomerase II inhibition. Strong catalytic inhibition of topoisomerase II by liriodenine was confirmed by in vitro assays with purified human topoisomerase II and kinetoplast DNA. Liriodenine also caused low-level protein-DNA cross-links to pulse-labeled SV40 chromosomes in vivo, suggesting that it may be a weak topoisomerase II poison. This was supported by the finding that liriodenine caused topoisomerase II-DNA cross-links in an in vitro assay for topoisomerase II poisons. Verapamil did not increase either liriodenine-induced protein-DNA cross-links or catalytic inhibition of topoisomerase II in SV40-infected cells. This indicates that liriodenine is not a substrate for the verapamil-sensitive drug efflux pump in CV-1 cells.

Topoisomerase II inhibition by aporphine alkaloids.

The aporphine alkaloids (+)-dicentrine and (+)-bulbocapnine are non-planar molecules lacking features normally associated with DNA binding by intercalation or minor groove binding. Surprisingly, dicentrine showed significant activity as a topoisomerase II (EC 5.99.1.3) inhibitor and also was active in a DNA unwinding assay. The DNA unwinding suggests DNA intercalation, which could explain the inhibition of topoisomerase II. Bulbocapnine, which differs from dicentrine only by the presence of a hydroxyl group at position 11 and the absence of a methoxyl group at position 9, was inactive in all assays. Molecular modeling showed that dicentrine can attain a relatively planar conformation, whereas bulbocapnine cannot, due to steric interaction between the 11-hydroxyl group and an oxygen of the methylenedioxy ring. These observations suggest that dicentrine is an "adaptive" DNA intercalator, which can bind DNA only by adopting a somewhat strained planar conformation. The requirement of a suboptimal conformation to achieve DNA binding appears to make dicentrine a weaker topoisomerase II inhibitor than the very planar oxoaporphine alkaloid liriodenine. These results suggest that it may be possible to modulate DNA binding and biologic activity of drugs by modifications affecting their ability to adopt planar conformations.

Gene amplification as a target for cancer chemotherapy.

Facile gene amplification is one aspect of the genetic instability associated with transformed cells. Amplification of oncogenes and proto-oncogenes contributes to carcinogenesis and tumor progression. Gene amplification is also a common basis for resistance to anticancer drugs. The observation that low level cytotoxic stress can cause rapid loss of amplified genes from cultured cell populations suggests that gene amplification may be a potential target for cancer chemotherapy. Drug-induced loss of amplified genes is seen with a wide variety of extrachromosomally amplified genes, including drug resistance genes and proto-oncogenes. A number of drugs and differentiating agents have been reported to cause rapid loss of unstably amplified genes. An effect on amplified genes or cells carrying amplified genes may contribute to the selective action of drugs presently used for cancer chemotherapy. A better understanding of drug-induced amplified gene loss may lead to new strategies for cancer treatment.

Topoisomerase inhibitors can selectively interfere with different stages of simian virus 40 DNA replication.

I have found that type I and type II topoisomerase inhibitors have pronounced but very different effects on SV40 DNA replication. Type II topoisomerase inhibitors selectively interfere with separation of newly replicated SV40 daughter chromosomes, while the type I topoisomerase inhibitor camptothecin rapidly produces broken replication forks in growing Cairns structures. Camptothecin also produces abnormal SV40 DNAs which appear to be intermediates and end products of sister chromatid exchange.

SV40 DNA replication intermediates: analysis of drugs which target mammalian DNA replication.

The simian virus 40 chromosome, a model for the mammalian replicon, is a uniquely powerful system for the study of drugs and treatments which target enzymes of the mammalian replication apparatus. High resolution gel electrophoretic analysis of normal and aberrant viral replication intermediates can be used effectively to understand the molecular events of replication failure. These events include breakage of replication forks, aberrant topoisomerase action, failure to separate daughter chromosomes, protein-DNA crosslinking, single and double strand DNA breakage, alterations in topology and inactivation of replication intermediates. The SV40 replication system can also be used to study the recombinational events which often follow drug-induced replication failure.

Methotrexate resistance in NIH3T3 cells expressing polyoma virus oncogenes.

NIH3T3 cell lines expressing different polyoma virus oncogenes were found to differ by almost three orders of magnitude in frequencies of methotrexate-resistant (MTXr) cells in single-step selections. Cells expressing the polyoma small T antigen showed the highest frequencies of MTX resistance, whereas cells expressing the polyoma middle T antigen showed the lowest frequencies of MTX resistance. Amplification of the dihydrofolate reductase (dhfr) gene was not a mechanism of MTX resistance in these cell lines. Cells expressing large T antigen also showed a high frequency of MTXr colonies with only two of 20 such colonies showing evidence of marginal (less than threefold) dhfr gene amplification. Cells carrying the cloning vector without an oncogene showed intermediate frequencies of MTX resistance, and in contrast to the oncogene expressing lines, two-thirds of these MTXr colonies showed marked dhfr gene amplification. The results suggest that the plasticity of the genome as measured by frequency or extent of dhfr gene amplification is not increased by expression of either immortalizing or transforming polyoma oncogenes. These oncogenes may, however, influence MTX resistance through epigenetic mechanisms. In addition, there appears to be no relationship between frequency of methotrexate resistance and frequency of dhfr gene amplification. It is clear that mechanisms of resistance other than dhfr gene amplification can be exclusively associated with either very high or very low frequencies of resistance.

Loss of unstably amplified dihydrofolate reductase genes from mouse cells is greatly accelerated by hydroxyurea.

Previous work has shown that mammalian cells that carry unstably amplified genes for dihydrofolate reductase (DHFR) gradually lose the amplified DHFR genes when grown in the absence of the DHFR inhibitor methotrexate (MTX). Unstably amplified genes occur on small acentric chromosomes called double minutes (DMs) or even smaller chromatin fragments, in contrast to stably amplified genes, which reside in centromere-containing chromosomes. We have found that the rate of loss of the unstably amplified DHFR genes can be greatly oncreased by growing the cells in the presence of a nonlethal concentration of hydroxyurea. For example, in one MTX-resistant subline studied, approximately equal to 90% of the original DHFR gene dosage is lost in 25-30 cell doublings in the absence of MTX. The same degree of loss is achieved, however, in less than 4 doublings if cells are grown in the presence of 50 microM hydroxyurea. This new effect of hydroxyurea does not appear to be due to changes in plating efficiency or selective cytotoxicity. In particular, no increase in cell death occurs at 50 microM hydroxyurea, and cells continue to multiply, albeit 1/2 to 2/3 as fast as in the absence of hydroxyurea. The ability to selectively accelerate the loss of amplified genes from mammalian cells as shown in the present work may have important implications both for the problem of drug resistance in cancer chemotherapy and for curing mammalian cells of extrachromosomally maintained DNA genomes of pathogenic viruses.

Escherichia coli photoreactivating enzyme: purification and properties.

We have purified large quantities of Escherichia coli photoreactivating enzyme (EC 4.1.99.3) to apparent homogeneity and have studied its physical and chemical properties. The enzyme has a molecular weight of 36 800 and a S020,W of 3.72 S. Amino acid analysis revealed an apparent absence of tryptophan, a low content of aromatic residues, and the presence of no unusual amino acids. The N terminus is arginine. The purified enzyme contained up to 13% carbohydrate by weight. The carbohydrate was composed of mannose, galactose, glucose, and N-acetylglucosamine. The enzyme is also associated with RNA (approximately 10 nucleotides/enzyme molecule) containing uracil, adenine, guanine, and cytosine with no unusual bases detected.

Efficiency of formation of pyrimidine dimers in SV40 chromatin in vitro.

The efficiency of formation of pyrimidine dimers by 254-nm light was studied in mixtures of SV40 chromatin and DNA extracted from that chromatin. At high doses (beyond 380 J/m2), fewer dimers are formed in chromatin than in DNA for a given dose of radiation. This difference is about 10% as saturation with pyrimidine dimers is approached at 6840 J/m2. Conversely, at biologically repairable doses (up to 40 J/m2, less than 2 dimers/genome), significantly more dimers are produced in the chromatin than in the DNA. A maximum increase of about 50% occurs at doses producing 0.5--20 dimers/genome. With isolated nucleosomes from this chromatin, a maximum increase in dimer formation of 77% was observed. Therefore, the increased dimer formation in the whole chromatin can be wholly accounted for in the nucleosome portion.

Patterns of strongly protein-associated simian virus 40 DNA replication intermediates resulting from exposures to specific topoisomerase poisons.

Exposure of infected CV-1 cells to specific type I and type II topoisomerase poisons caused strong protein association with distinct subsets of simian virus 40 (SV40) DNA replication intermediates. On the basis of the known specificity and mechanisms of action of these drugs, the proteins involved are assumed to be the respective topoisomerases. Camptothecin, a topoisomerase I poison, caused strong protein association with form II (relaxed circular) and form III (linear) viral genomes and replication intermediates having broken DNA replication forks but not with form I (superhelical) viral DNA or normal late replication intermediates which were present. In contrast, type II topoisomerase poisons caused completely replicated forms and late viral replication forms to be tightly bound to protein--some to a greater extent than others. Different type II topoisomerase inhibitors caused distinctive patterns of protein association with the replication intermediates present. Both intercalating and nonintercalating type II topoisomerase poisons caused a small amount of form I (superhelical) SV40 DNA to be protein-associated in vivo. The protein complex with form I viral DNA was entirely drug-dependent and strong, but apparently noncovalent. The protein associated with form I DNA may represent a drug-stabilized "topological complex" between type II topoisomerase and SV40 DNA.

Drug-induced loss of unstably amplified genes.

Methotrexate-resistant R500 cells slowly lose amplified dihydrofolate reductase (dhrf) genes with biphasic kinetics when grown in the absence of methotrexate. Both phases of gene loss were markedly accelerated by subcytotoxic drug treatments. R500 cells were passed in low concentrations of cytotoxic drugs (inhibitors of ribonucleotide reductase, type I and type II topoisomerases, and polyamine synthesis). At each passage, relative dhfr gene copy number was determined by slot blot analysis. All of these drugs were able to induce rapid loss of dhfr gene dosage in the R500 cell population. The ability of these treatments to cause the rapid emergence of a cell population with substantially reduced dhfr gene dosage indicates that either the amplified genes or those cells with the highest levels of gene amplification are selectively targeted by low-level cytotoxic stress. The complex kinetics of amplified gene loss are suggestive of differential targeting of resistant cell subpopulations.

Exposure to camptothecin breaks leading and lagging strand simian virus 40 DNA replication forks.

To better understand aberrant simian virus 40 DNA replication intermediates produced by exposure of infected cells to the anticancer drug camptothecin, we compared them to forms produced by S1 nuclease digestion of normal viral replication intermediates. All of the major forms were identical in both cases. Thus the aberrant viral replicating forms in camptothecin-treated cells result from DNA strand breaks at replication forks. Linear simian virus 40 forms which are produced by camptothecin exposure during viral replication were identified as detached DNA replication bubbles. This indicates that double strand DNA breaks caused by camptothecin-topoisomerase I complexes occur at both leading and lagging strand replication forks in vivo.

Aldehyde-induced protein-DNA crosslinks disrupt specific stages of SV40 DNA replication.

Aldehydes with specific protein-DNA crosslinking ability disrupted simian virus 40 (SV40) DNA replication to cause replication fork failure by the 40S intermediate pathway, in which replicating viral genomes become inactivated and torsionally stressed. In contrast, aldehydes without detectable protein-DNA crosslinking ability had no effect on SV40 DNA replication during the 10 min exposure times employed. This indicates that protein-DNA crosslinks block either DNA polymerase or the entire replication complex. Replication failure by the 40S pathway is known to initiate recombinational events in the damaged SV40 replicons. Similar events in cellular replicons may play a role in the clastogenic effects of formaldehyde. In addition, formaldehyde and acrolein caused accumulation of catenated (topologically linked) SV40 daughter chromosomes--a signature of topoisomerase II inhibition.