Genomic sites of topoisomerase II activity determined by comparing DNA breakage enhanced by three distinct poisons¹

Abstract

To define the sites of topoisomerase II activity in two genomic regions of Drosophila melanogaster Kc cells, we have investigated in vivo DNA cleavage sites stimulated by three poisons with diverse sequence specificity, clerocidin, VM-26 and dh-EPI (an anthracycline analog). DNA cleavage was studied by PFGE (pulse-field gel electrophoresis), standard gel electrophoresis, and by genomic primer extension. Poisons stimulated specific intensity patterns of cleavage in the two genomic regions studied. At the centromeric satellite III DNA, fragments of about 270–310 and 385–430 kb could be detected specifically after treatment with clerocidin, suggesting a complex DNA loop organisation, which may correspond with a centromere-specific higher-order chromatin structure. Clerocidin-dependent DNA fragmentation was detectable by PFGE but not by standard agarose gel electrophoresis; while VM-26-dependent cleavage was detected with either method, dh-EPI was ineffective at this locus. Thus, clerocidin DNA cleavage sites were rarer than those of VM-26 at the satellite locus. In the histone H2A-H2B intergenic region, clerocidin and dh-EPI stimulated cleavage whereas VM-26 was only weakly effective. Some clerocidin cleavage sites did not undergo spontaneous reversion, indicating that this agent can stimulate irreversible cleavage in vivo. Direct genomic sequencing showed that many clerocidin and dh-EPI sites, although distinct, mapped to the transcription start and to the proximal promoter of the H2A gene, suggesting that the region is highly accessible to topoisomerase II. Thus, the enzyme may play a role in maintaining a highly accessible chromatin structure under normal cell growth conditions, possibly mediated by specialised protein complexes. This study demonstrates that the use of distinct poisons greatly improves the definition of genomic sites of topoisomerase II activity.

Introduction

DNA topoisomerases are essential enzymes that modify the topology of nucleic acids, and can be classified into distinct families based on sequence homology and conserved catalytic mechanism (Wang, 1996). Type II DNA topoisomerases introduce double-stranded breaks into DNA segments, transport a second DNA segment through the cut, and finally reseal cut strands. Biochemical and genetic data demonstrate that DNA topoisomerase II is essential for the segregation of newly replicated pairs of intertwined chromosomes Holm et al 1985, Downes et al 1991, Adachi et al 1991, and is involved in the modulation of the supercoiling state of intracellular DNA (Liu & Wang, 1987). Eukaryotic topoisomerases are targets of very effective antitumor drugs that interfere with the enzyme activity by stabilising a reaction intermediate wherein DNA strands are cut and covalently linked to tyrosine residues of the protein Pommier 1997, Froelich-Ammon and Osheroff 1995, Capranico et al 1997, Wang 1996. The enzyme-poisoning action of natural compounds has been conserved during evolution, most likely because it is a very efficient mechanism of cell killing.

Topoisomerase II is a structural component of the nuclear matrix of interphase cells, is a major component of chromosome scaffolds, and is required for mitotic chromosome condensation (Poljak & Käs, 1995). Moreover, specific DNA helicases likely associate with topoisomerase II for faithful transmission of chromosomes to daughter cells Watt et al 1995, Watt et al 1996. Recently, a multi-protein complex (CHRAC) containing topoisomerase II has been shown to remodel nucleosome arrays and to increase the accessibility of DNA to other protein factors, thus suggesting a role of CHRAC in the unravelling of repressed chromatin domains (Varga-Weisz et al., 1997). Recent reports emphasise the idea that nuclear enzyme molecules may constitute diverse pools of functionally distinct topoisomerase II proteins. The enzyme is distributed widely over interphase chromatin, but is also concentrated in specific regions of heterochromatin (Swedlow et al., 1993). In addition, distinct populations of topoisomerase II have been distinguished by different dynamic patterns in chromosomes (Swedlow et al., 1993).

To achieve a more complete understanding of topoisomerase II functions, much attention has focused on the sites of topoisomerase II activity in the chromatin of living cells. This approach has been made possible by the use of antitumor poisons to increase the level of DNA cleavage Rowe et al 1986, Yang et al 1985, Udvardy et al 1985, Reitman and Felsenfeld 1990, Udvardy and Schedl 1991, Kas and Laemmli 1992, Gromova et al 1995b, Borde and Duguet 1996, Ebert et al 1990, Pommier et al 1992. Topoisomerase II has thus been located at scaffold-attachment regions (SARs) Gasser and Laemmli 1986, Adachi et al 1989, Poljak and Kas 1995 and at domain boundary elements (Udvardy & Schedl, 1993), suggesting a role in chromatin organisation. The size of chromatin loops has also been estimated through the use of topoisomerase II poisons as revealed by the length of DNA fragments released from drug-treated nuclei Gromova et al 1995a, Iarovaia et al 1996. DNA cleavage sites were also found to co-localise with DNase I-hypersensitive sites at gene promoters and to be mainly excluded from gene coding sequences Udvardy et al 1985, Reitman and Felsenfeld 1990, Udvardy and Schedl 1991, Kas and Laemmli 1992.

In this approach topoisomerase II poisons are used as a tool to localise the physiological sites of enzyme activity in chromatin domains. Thus, the molecular mechanism of poison interference with the enzyme is of great importance not only for the development of novel effective anticancer drugs, but also for the full understanding of enzyme functions. The molecular action of poisons is peculiar; these agents likely bind to a protein-DNA interface at the enzyme active site, thus hindering strand religation activity of topoisomerases (Capranico et al., 1997). This view is based on several lines of evidence, including sequence specificity of poison action Pommier et al 1993, Capranico et al 1997, photolabelling of DNA by cross-reactive poison analogs Freudenreich and Kreuzer 1994, Pommier et al 1995, and drug binding data Shen et al 1989, Hertzberg et al 1989, Hertzberg et al 1990. In the case of camptothecin, a topoisomerase I poison, models for the drug receptor have been proposed based on the crystal structures of enzyme-DNA complexes (Redinbo et al., 1998) and molecular modelling analyses (Fan et al., 1998). Camptothecin contacts specific amino acid residues as well as the G·C base-pair at the +1 position of the cleavage site that was previously shown to be required for its activity (Pommier et al., 1993).

A single poison (VM-26, 4′-demethylepipodophillotoxin-9-(4,6-O-thionylidine-β-d-glucopyrano-side) has often been used as a tool to map topoisomerase II activity in chromatin domains by low-resolution agarose gel electrophoresis. Since available data demonstrate that VM-26 maintains in vivo the sequence specificity established in vitro(Borgnetto et al., 1996), these investigations are inherently limited. The use of a single poison can indeed fail to reveal genomic sites of enzyme activity where the poison is inefficient in stimulating cleavage (Borgnetto et al., 1996). Thus, in the present work, we have investigated topoisomerase II-dependent DNA cleavage stimulated by clerocidin, VM-26 and dh-EPI (4′-demethyory-3′-deamino-3′-hydroxy-4′-epi doxorubicin), a potent anthracycline analog, in Drosophila Kc cells. The three poisons have markedly different sequence specificity in vitro requiring a guanine, a cytosine and a 5′-TA dinucleotide, respectively, at the 3′ terminus generated by topoisomerase II Capranico et al 1994, Capranico et al 1997. We have shown that clerocidin, unlike the other poisons, mainly stimulated irreversible DNA cleavage in SV40 DNA (Binaschi et al., 1997), since it likely forms a covalent adduct with the 3′ terminus guanine. The studied genomic regions of Drosophila Kc cells were known to be sites of enzyme activity, and topoisomerase II cleavage sites were in part sequenced in these regions in the case of VM-26 and dh-EPI Kas and Laemmli 1992, Borgnetto et al 1996. Our present findings demonstrate that the use of distinct poisons greatly improves the definition of genomic sites of topoisomerase II activity. Indeed, we present evidence that topoisomerase II may be involved in maintaining diverse chromatin structures at the two genomic loci studied.