Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Eukaryotic topoisomerase II (topo II) is a gene involved in essential cellular processes including chromosomal segregation at mitosis. In human cells, there are two expressed isoforms of topo II, termed  and , each located on a distinct genetic locus. The topo II form has a molecular mass of 170 kDa, maps to chromosome 17q21-22 (Wang, 1996); expression is increased in proliferating cells and correlates with cellular S-phase fraction (Holden et al., 1992; Sandri et al., 1996a). There is increased expression of topo II in tumor cells compared with nonmalignant tissue (Kim et al., 1991; Hasegawa et al., 1993). However, there is marked heterogeneity of topo II expression within a population of cancer cells (Turley et al., 1997). Furthermore, there is evidence that several oncogenes, including ras (Chen et al., 1999), myb (Brandt et al., 1997), and p53 (Sandri et al., 1996b; Brandt et al., 1997) interact with sequences within the topo II promoter. Topo II is the target of several important anticancer agents such as doxorubicin and etoposide. The enzyme can pass an intact DNA helix through a transient double-stranded break that it generates in a separate strand (Burden and Osheroff, 1998). Topo II poisons can stabilize the cleavable complex, resulting in an increase in double-stranded breaks that ultimately results in cell death.

Several studies (Davies et al., 1988; Fry et al., 1991) have indicated that up-regulation of topo II within tumor cells may result in increased sensitivity to doxorubicin and etoposide. The transcriptional expression of topo II is reduced under conditions of cellular confluence (Sullivan et al., 1987). The reduction in topo II levels probably contributes to the increased resistance to topo II poisons in confluent cells. There is increasing evidence that experimental models using cells cultured under conditions of confluence may be better models than logarithmically growing cultures (Desoize and Jardillier, 2000; Padrón et al., 2000). Although depletion of growth factors does not seem to be a significant factor in the reduction of topo II levels at confluence (Isaacs et al., 1996), intercellular contact in confluent models may more closely approximate the situation within tumors than cells in exponential phase of growth.

The sequences within the topo II promoter influencing confluence-induced down-regulation have been identified. Using a series of stable transfectants, including the topo II promoter linked to a growth hormone reporter in the Swiss 3T3 cell line, the inverted CCAAT box 2 (ICB2) at position 108 to 104 relative to the transcription start site was shown to be required for confluence induced down-regulation (Isaacs et al., 1996). Experiments using a gel shift assay identified NF-Y, a heterotrimeric transcription factor, as a component of the complex binding this sequence. There is reduced expression of topo II in areas of confluence and hypoxia (Tomida and Tsuruo, 1999) and this may decrease the therapeutic efficacy of topo II poisons. The significance of topo II as a target for several important anticancer agents such as etoposide and doxorubicin further increases the importance of understanding factors affecting the regulation of this gene.

This study aimed to analyze the protein-DNA interactions responsible for transcriptional repression at confluence and investigate strategies to modulate topo II expression. We show that DNA-binding drugs with affinity for adenine-thymine (AT)-rich sequences are able to bind these elements and thereby modulate topo II transcription. In addition, Hoechst 33342 induced up-regulation of topo II increases sensitivity of confluent-phase cells to etoposide.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Culture and Drug Treatment. The NIH3T3 murine cell line was cultured at 37°C under an atmosphere of 10% CO₂ in Dulbecco's modified Eagle medium (Autogen Bioclear, Calne, Wiltshire, UK) supplemented with 10% newborn calf serum (Life Technologies, Paisley, Scotland), 4.5 g/l glucose and 2 mM L-glutamine. The human keratinocyte K1, CaCo2 colon cancer, and HeLa cell lines were grown as the NIH3T3 line, except 10% fetal bovine serum (Autogen Bioclear) and 5% CO₂ atmosphere were used. Drug treatment was carried out with exponential growing NIH3T3 cells or cells that were kept for 48 h at confluence. Hoechst 33258 or 33342 (Sigma, St. Louis, MO) at various concentrations or no drug as a control, was added in fresh, prewarmed media and the cells were subsequently incubated for 24 h at 37°C. AR-1-144 was kindly donated by Dr. Moses Lee (Dept. of Chemistry, Furman University, Greenville, SC) and distamycin A was obtained from Sigma.

Cytotoxicity Assays. Exponential or confluent phase cells were incubated for 8 h with Hoechst 33342, washed with prewarmed (37°C) complete medium, and subsequently incubated overnight with etoposide (Sigma). The treated cells were washed with prewarmed (37°C) complete medium and incubated for a further 36 h. At this point, 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazoliumbromide (MTT; Sigma) was added to 0.5 mg/ml and, after a further 2-h incubation, the medium was removed, the cells were lysed, and the dye dissolved by adding dimethyl sulfoxide (BDH, Poole, Dorset, UK). The reduction of MTT was quantified by measuring the absorbance at 540 nm.

Reporter Constructs, Site Directed Mutagenesis, and Transient Transfections. To determine the effect of Hoechst 33342 and the dominant negative NF-YA mutant NF-Y29 (construct pNF-Y29; Mantovani et al., 1994) on topo II promoter activity, the 557-bp promoter fragment of pCAT557 (Hochhauser et al., 1992) was subcloned upstream of the promoterless luciferase reporter gene in pGL3-basic (Promega, Madison, WI), designated pT2WT. A topo II construct carrying a polymerase chain reaction-generated mutation in ICB2 (primers ICB2 MFW, sense, 5'-GGCAAGCTACGTTTCCTTCTTCTGGACG-3'; ICB2 MAS, antisense, 5'-CGTCCAGAAGAAGGAAACGTAGCTTGCC-3') but otherwise identical to pT2WT was designated pT2 MT. The mutation in ICB2 was verified with a BESS Mutascan mutation characterization kit as described by the manufacturer (Epicentre Technologies, Madison, WI). Subsequently, the insert of the final construct was single stranded sequenced using the dideoxy termination method using an automated sequencing system (MWG-Biotech) and compared with the wild-type promoter sequence. The pGL3-promoter vector (Promega) was used to study the effects of Hoechst 33342 on a topo II unrelated promoter (i.e., without an inverted CCAAT box).

Preparation of Nuclear Extracts. Nuclear extracts were essentially prepared as described previously (Firth et al., 1994) and all steps were performed at 4°C in the presence of a protease inhibitor mix (Complete; Roche Molecular Biochemicals, Mannheim, Germany). Briefly, cells were rinsed with ice-cold phosphate-buffered saline, scraped from the surface, and collected by centrifugation. The cells were washed with 5 equivolumes of hypotonic buffer containing 10 mM K-HEPES, pH 7.9, 1.5 mM MgCl₂, 10 mM KCl, 0.5 mM dithiothreitol (DTT; Sigma). Subsequently, the cells were resuspended in 3 equivolumes of hypotonic buffer, incubated on ice for 10 min, subjected to 20 strokes of a Dounce homogenizer, and the nuclei were collected by centrifugation. The nuclear pellet was resuspended in 0.5 equivolumes of low-salt buffer containing 20 mM K-HEPES, pH 7.9, 0.2 mM K-EDTA, 25% glycerol, 1.5 mM MgCl₂, 20 mM KCl, and 0.5 mM DTT. While stirring, 0.5 equivolumes of high-salt buffer (as low-salt buffer, but containing 1.4 M KCl) was added and the nuclei were extracted for 30 min. Subsequently, the mixture was centrifuged for 30 min at 14,000 rpm in an Eppendorf centrifuge and the supernatant was dialysed in tubing with a 12-kDa cut off (Sigma) for 1 h in a 100× excess of dialysis buffer containing 20 mM K-HEPES, pH 7.9, 0.2 mM K-EDTA, 20% glycerol, 100 mM KCl, and 0.5 mM DTT. The dialysed fraction was centrifuged for 30 min at 14,000 rpm in an Eppendorf centrifuge and the supernatant was snap frozen in an ethanol dry ice bath and stored at 80°C. The protein concentration of the nuclear extract was assayed using a Bio-Rad micro protein assay kit (Bio-Rad, Hercules, CA).

Western Blot Analysis. For Western blot analysis, 5 µg of nuclear extract was denaturated by heating for 3 min at 95°C in sample buffer containing 100 mM Tris-Cl, pH 6.8, 4% SDS, 10% 2-mercaptoethanol, 20% glycerol, and 0.02% bromphenol blue (BFB). Bio-Rad high range SDS-polyacrylamide gel electrophoresis molecular mass standards were used as a reference. Proteins were separated on a 7% SDS-polyacrylamide mini gel (Mini Protean II system; Bio-Rad) and subsequently transferred (Trans Blot Cell; Bio-Rad) to polyvinylidene difluoride membranes (Immobilon-P; Millipore, Bedford, MA). Western blot analysis was performed with the IHIC8 rabbit polyclonal topoisomerase II antibody (kindly provided by Dr. I. D. Hickson, Institute for Molecular Medicine, Oxford, UK) at a 1:5000 dilution using a enhanced chemiluminescence Western blot detection kit and protocol (Amersham Pharmacia Biotech, Piscataway, NJ) using 1% blot qualified bovine serum albumin (Promega) as blocking reagents and Tris-buffered saline plus 0.5% Tween 20 (BDH) as a buffer. The chemiluminescent signal was visualized by exposing the blots to Kodak X-Omat-LS film (Kodak, Rochester, NY).

Electrophoretic Mobility Shift Assay (EMSA). The following oligonucleotides (Genosys, Cambridgeshire, UK) containing ICBs (underlined) were used in EMSAs: ICB1 sense, 5'-CGAGTCAGGGATTGGCTGGTCTGCTTC-3'; ICB1 antisense, 5'-GAAGCAGACCAGCCAATCCCTGACTCG-3'; ICB2 sense, 5'-GGCAAGCTAC GATTGGTTCTTCTGGACG-3'; ICB2 antisense, 5'-CGTCCAGAAGAACC AATCGTAGCTTGCC-3'; ICB3 sense, 5'-CTCCCTAACCTGATTGGTTTATT CAAAC-3'; ICB3 antisense, 5'-GTTTGAATAAACCAATCAGGTTAGGGAG-3'; ICB4 sense, 5'-GAGCCCTTCTCATTGGCCAGATTCCCTG-3'; ICB4 antisense, 5'-CAGGGAATCTGGCCAATGAGAAGGGCTC-3'; ICB5 sense, 5'-GATCTTAAATAGATTGGCAGTTCCTGGAG-3'; ICB5 antisense, 5'-CTCCAG GAACTGCCAATCTATTTAAGATC-3'. Oligonucleotides containing mutated ICBs were used as specific competitors of similar sequence, except the wild-type ICB sequence was replaced by AAACC or GGTTT in sense and antisense oligonucleotides, respectively. Sense and antisense oligonucleotides were annealled in an equimolar ratio. Double-stranded oligonucleotides were 5' end labeled with T4 kinase (New England Biolabs, Hitchin, Hertfordshire, UK) using [-³²P]ATP and subsequently purified on Bio-Gel P-6 columns (Bio-Rad). EMSAs were essentially performed as described previously (Firth et al., 1994). Briefly, 5 µg of nuclear extract in a total volume of 10 µl was incubated at 4°C for 30 min in a buffer containing 20 mM K-HEPES pH 7.9, 1 mM MgCl₂, 0.5 mM K-EDTA, 10% glycerol, 50 mM KCl, 0.5 mM DTT, and 0.5 µg poly(dI-dC). poly(dI-dC) (Pharmacia) and 1× protease inhibitor mix (Complete; Roche Molecular Biochemicals). For supershifts, antibodies against NF-YA or NF-YB (IgG fraction; Rockland, Gilbertsville, PA) were used and the preincubation on ice was extended for a total of 1.5 h. Upon addition of approximately 0.1 ng of radiolabeled probe, the incubation was continued for 30 min at room temperature. In competition experiments, radiolabeled probe and competitor were added simultaneously. Subsequently, 0.5 µl of loading buffer (25 mM Tris-Cl, pH 7.5, 0.02% BFB and 10% glycerol) was added and the samples were separated on a 4% polyacrylamide gel in 0.5× Tris/borate/EDTA buffer containing 2.5% glycerol at 4°C. After drying the gels, the radioactive signal was visualized by exposing the gels to Kodak X-Omat-LS film.

DNase I Footprinting. A radiolabeled probe of 479 bp corresponding to positions 489 through 10 relative to the transcriptional start site of the topo II promoter was generated as follows. Antisense oligonucleotide (4 pmol) 5'-GTCGGTTAGGAGAGCTCCACTTG-3' was 5' end-labeled with T4 kinase (New England Biolabs) using [-³²P]ATP in a 10-µl reaction, followed by heat inactivation for 20 min at 65°C. Subsequently, 4 pmol of sense oligonucleotide (5'-CTGTCCAGAAAGCCG GCACTCAG-3'), 2 µl of 10 mM dNTPs (Promega), 1 U of Red Hot DNA Polymerase (Abgene, Epsom, Surrey, UK), 2 µl of 25 mM MgCl₂, and 4.5 µl of 10× reaction buffer IV (Abgene) were added (in a final volume of 50 µl) and a polymerase chain reaction was performed consisting of: 3 min at 95°C and 1 min at 95°C, 1 min at 60°C, and 2 min at 72°C for 35 cycles. The product was purified on a Bio-Gel P-6 column (Bio-Rad). DNase I footprint reactions were performed with 30 µg of nuclear extract in a 50-µl reaction in the same buffer as used for EMSA. After preincubation for 30 min at 4°C, approximately 0.1 ng of radiolabeled probe was added and the mixture was incubated at room temperature for another 30 min. Subsequently, 1 U of RQ1 DNase I (Promega) and up to 5 mM MgCl₂ and CaCl₂ were added. After exactly 3 min of digestion at room temperature, 1 volume of stop mix containing 30 mM K-EDTA, pH 8.0, 200 mM NaCl, and 1% SDS was added and samples were purified by phenol-chloroform treatment and alcohol precipitation. The resulting pellets were dried and resuspended in loading buffer (95% formamide, 20 mM K-EDTA, pH 8.0, 0.05% BFB, and 0.05% xylene cyanol). The sample was heat-denaturated for 3 min at 95°C and separated on a 6% denaturating polyacrylamide gel (Sequagel; National Diagnostics, Hessle Hull, UK). A 10-bp ladder (Life Technologies) labeled with ³²P by T4 kinase was used as a molecular mass standard. The dried gels were exposed to Kodak X-Omat-LS film with intensifying screens (Kodak) at 80°C.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Investigation of Effects of Confluence on the Topo II Promoter. To verify down-regulation of topo II at confluence in NIH3T3 cells, nuclear extracts from exponential and confluent phase cells were immunoblotted with an antibody for topo II (Fig. 1a). As previously demonstrated, there is dramatic down-regulation of topo II expression at confluence; this has been shown to be caused by alterations in transcriptional regulation (Isaacs et al., 1996). Similar patterns were seen in extracts from the human K1 keratinocyte and CaCo2 colon cancer cell lines in which down-regulation of topo II at confluence was also demonstrated. In contrast, the cervical cancer cell line HeLa showed little alteration of topo II expression in confluent cells.

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Down-regulation of topo II at confluence and organization of the human topo II promoter. a, Western blot analysis of nuclear extracts isolated from NIH3T3, Hela, CaCo2, and K1 cells. The blot was probed with antibodies against topo II. E, exponential phase; C, confluent phase. b, relative position of putative cis-acting elements in the topo II promoter. GC, GC box; ICB1-5, inverted CCAAT boxes 1 to 5; ATF, activating transcription factor binding site; ATG, topo II translation start codon; HSE, heat-shock element; shaded box, ICB involved in topo II down-regulation at confluence.

View larger version (33K): [in this window] [in a new window]   Fig. 2.   EMSA, DNase I footprinting and luciferase reporter assay showing involvement of NF-Y in the complex that binds to ICB2. a, nuclear extracts prepared from NIH3T3 cells in exponential and confluent phase (lanes 1-6) were incubated with a radioactive oligonucleotide containing the ICB2 sequence of the topo II promoter. Lane 1, control lane containing probe and nuclear extracts only. Lanes 2 and 3 contain an unlabeled competitor oligonucleotide containing the wild-type or mutated NF-YA binding site, respectively. Lanes 4, 5, and 6 contain NF-YA, NF-YB, and (as a control) Sp1 antibodies, respectively. F, free probe; S, specific DNA/protein complex; Su, super shifted DNA/protein complex; Exp, nuclear extracts of exponential phase cells; Con, nuclear extracts of confluent phase cells. b, DNase I footprint of nuclear extracts isolated from NIH3T3 cells in exponential (lane 1) and confluent phase (lane 2). The position of ICB1 and 2 is indicated relative to the CCAAT sequence position (bars) and the actual footprint (bracket). M, molecular mass marker (increments of 10 bp); and C, control (no protein). c, DNase I footprint of nuclear extracts isolated from NIH3T3 cells in confluent phase with a probe containing the wild-type (lane 1) or mutated (lane 2) ICB2 sequence. The position of ICB1 and 2 is indicated relative to the CCAAT sequence position (bars) and the actual footprint (bracket). M, molecular mass marker (increments of 10 bp); C1, control lane (no protein) wild-type ICB2 sequence; C2, control lane (no protein) mutant ICB2 sequence. d, effect of dominant negative NF-YA on topo II promoter activity. Luciferase reporter assays were performed on extracts isolated from exponential (lanes 1, 4) and confluent (lanes 2, 3, 5-8) phase NIH3T3 cells. The combination of plasmids used is indicated below the figure. PT2WT, wild-type topo II promoter; pT2 MT, topo II promoter with mutated ICB2; pGL3-basic, empty vector control for pT2WT and pT2 MT; pNF-Y29, dominant negative NF-YA expression vector; pCDNA3, empty vector control of pNF-Y29.

Effects of DNA Minor Groove Binding Drugs on the Topo II Promoter. Several DNA minor groove binding drugs (MGB) have been shown to bind preferentially to AT rich sequences in DNA. These include the oligopeptide antibiotic distamycin and the bis-benzimidazoles Hoechst 33342 and Hoechst 33258 (Nielsen, 1991). In view of the ubiquitous presence of CCAAT sequences within the topo II promoter region, EMSA experiments were carried out using labeled oligonucleotide probes containing sequences of the ICB2 found in the promoter (Fig. 3). The probes were incubated with cell extracts and increasing concentration of drug. There was inhibition of protein complex binding to the ICB sequence with increasing drug concentration for the MGB Hoechst 33342 and Hoechst 33258 (Fig. 3, a and b). In contrast, addition of Hoechst 33342 and Hoechst 33258 had no effect on a radiolabeled oligonucleotide containing the GC sequences implicated in Sp3 binding (Mo et al., 1997) (Fig. 3c). With distamycin, Hoechst 33342 and Hoechst 33258 abolition of binding was seen at 30 µM. Incubation with the agent AR-1-44, an imidazole derivative of distamycin targeting guanine and cytosine rich sequences (Lee et al., 1993), results in only a partial inhibition at drug concentrations above 100 µM (Fig. 3d).

View larger version (51K): [in this window] [in a new window]   Fig. 3.   Inhibition of protein complex binding to ICB2 by AT specific minor groove binders. Effect of minor groove binders on EMSA of nuclear extracts isolated from NIH3T3 cells in exponential phase. Nuclear extracts were incubated with a radiolabeled probe containing the ICB2 wild-type binding site (a, b, d) or a GC-box sequence (c). In each case, lane 1 contains unlabeled competitor probe and lane 2 contains mutated probe. Lanes 3 to 7 contain increasing concentrations of minor groove binding agent as indicated at concentrations of 10, 20, 30, 40, and 50 µM (Hoechst 33342 and Hoechst 33258) or 20, 40, 60, 80, and 100 µM (AR-1-144).

View larger version (55K): [in this window] [in a new window]   Fig. 4.   Displacement of protein complexes from the topo II promoter by minor groove binders. Effect of distamycin A and Hoechst 33342 on DNase I footprinting in the absence (a, b) or presence (c) of nuclear extracts isolated from NIH3T3 cells in exponential phase. Extracts were incubated with MGBs and a radiolabeled probe of the topo II promoter. TATAA, nonfunctional TATAA-box; M, molecular mass marker (increments of 10 bp); C, control (no protein); solid and dashed brackets, Hoechst and nuclear extract footprints over the ICB2, respectively; bars, position of ICBs and TATAA-box relative to the promoter sequence; triangle, increasing amount of Hoechst 33342 (10, 20, 30, 40, and 50 µM).

Effect of Minor Groove Binding Drugs on Topo II Expression. The result of the EMSA and footprinting experiments indicated that incubation with Hoechst 33342 and Hoechst 33258 could prevent the binding of proteins inhibiting topo II expression. Furthermore, these results suggested that the affinity of these drugs for the CCAAT sequence is high and the pattern of binding does not alter with increasing drug concentration. To demonstrate the effect of drug incubation on topo II expression, NIH3T3 cells were grown to confluence and incubated with the various drugs. Exposure of cells to Hoechst 33342 resulted in a time- and concentration-dependent increased expression of topo II compared with cells grown in drug-free medium as demonstrated by immunoblotting (Fig. 5a). The increase in expression was to approximately the level of exponentially growing cells. A similar result was obtained with Hoechst 33258 (data not shown). This increased expression occurred within six h of treatment and was concentration-dependent (Fig. 5, b and c).

View larger version (19K): [in this window] [in a new window]   Fig. 5.   Increased expression of topo II in confluent cells following exposure to the minor groove binder Hoechst 33342. a, Western blot analysis of nuclear extracts isolated from exponential (lane 1), confluent (lane 2), and confluent phase cells treated with 100 µM Hoechst 33258 for 24 h (lane 3). b, time course of the incubation of confluent NIH3T3 cells with 100 µM Hoechst 33342. c, dose response of the incubation of confluent NIH3T3 cells with Hoechst 33342 for 6 h. The blots were probed with antibodies against topo II. d, luciferase reporter assays of nuclear extracts isolated from confluent phase NIH3T3 cells transfected with the topo II (pT2WT) or SV-40 promoter (pGL3). Cells were maintained at confluence for 48 h before adding the drug Hoechst 33342. The activity relative to pT2WT without drugs was set to 1. e, effect of Hoechst on the IC50 value of etoposide at confluence. MTT cytotoxicity assays were performed after successive incubation of confluent phase NIH3T3 cells with Hoechst 33342 for 8 h and etoposide for 16 h. The IC50 value of Hoechst 33342 alone in confluent phase cells was 50 µM.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

There is evidence that transcriptional control of topo II regulation is important in regulating gene expression at confluence (Isaacs et al., 1996), and under heat-shock conditions (Furukawa et al., 1998). A recent study indicated that transcriptional elements may regulate the cell cycle control of topo II expression (Falck et al., 1999). Furthermore, p53 may directly repress the promoter (Sandri et al., 1996b; Wang et al., 1997). The oncogenes ras and myb have also been demonstrated to act through specific sequences within the topo II promoter. Both heat-shock and cell cycle regulation have been demonstrated to act through the ICB1 in the promoter (Falck et al., 1999), whereas confluence-induced down-regulation is regulated through the ICB2. This study confirms that, unlike with cell cycle regulation, the transcription factor NF-Y forms part of the ICB binding complex. There is an alteration in the intensity of the in vitro footprint seen at confluence, although this is not associated with alteration in the amount of NF-YA or NF-YB expressed.

Previous studies have suggested that the inverted CCAAT boxes within the promoter are important in regulation of topo II transcription. Although NF-Y is significant in heat-shock response, a study on cell cycle regulation did not confirm NF-Y binding to the ICB1 site in the topo II promoter as being important in transcriptional repression (Falck et al., 1999). The effects of NF-Y are generally stimulatory on gene expression but NF-Y has also been shown to mediate inhibition of gene transcription of several genes including cdc2 (Yun et al., 1999). The role of other transcription factors in control of topo II expression has not been clarified. Although there is a clear footprint over the TATAA box the significance of this is unclear as the distance from the transcription initiation site suggests that it is nonfunctional (Hochhauser et al., 1992). Furthermore, expression of a novel 90-kDa CCAAT-box binding protein (ICBP90) that binds to the ICB2 consensus sequence coincides with that of topo II and is expressed in Hela cells at confluence (Hopfner et al., 2000).

The correlation between levels of topo II and chemosensitivity to topo II poisons has been shown in several studies, although other factors such as p53 status may also be important. Therefore, strategies to increase topo II expression may improve therapeutic effectiveness. The resistance to chemotherapeutic drugs seen in vitro at confluence is related in part to reduced expression of topo II (Dimanche-Boitrel et al., 1992), although altered membrane permeability to cytotoxic drugs may also play a role. It is unclear to what extent the reduced expression of topo II in tumors is caused by hypoxic down-regulation or by the confluence pathway investigated here. Further experiments with xenografts using animals treated with analogs of the agents used in this study should clarify this issue.

MGB agents have a variety of pharmacological effects. Hoechst 33258 binds to double-helical DNA (Bontemps et al., 1975) and has ready access into cells with minimal toxicity (Soderlind et al., 1999). It binds preferentially at AT-rich DNA sequences with a minimum of four consecutive AT base pairs. This may result in inhibition of transcription factor binding to DNA with consequent effect on gene regulation (Zewail-Foote and Hurley, 1999; White et al., 2000). These compounds have also been shown to act as topoisomerase I inhibitors (Chen et al., 1993). It has been demonstrated that exposure to Hoechst 33342 can directly alter gene expression by binding to promoter sequences within the c-fos promoter (White et al., 2000). Thus there are several mechanisms of action for these agents that are relatively sequence nonspecific; the development of novel lexitropsins will allow more precise sequence-specific targeting.

There is considerable potential to exploit the DNA sequence-specific properties of MGBs in modulating gene expression (Zewail-Foote and Hurley, 1999). The ability to overcome transcription factor-induced repression is important in the case of topo II because increased expression of the gene would increase sensitivity to topo II poisons. A component of topo II down-regulation in tumors may be caused by hypoxia, which has been shown to decrease topo II transcription, although the regions of the promoter modulating this have not been identified. However, it is likely that reduction in expression may also be modulated through the confluence pathway analyzed in this study, which may occur concurrently with hypoxia-induced repression. Abrogation of this repression may also induce apoptosis, which occurs after enforced expression of topo II (McPherson and Goldenberg, 1998). Other factors modulating topo II transcription have been identified. The trans-activation of topo II by c-Myb may account for the bulk of topo II promoter activity in human leukemia cells (Brandt et al., 1997). Furthermore, the Sp3 transcription factor is increased in cell lines resistant to topo II poisons with reduced transcription of topo II (Mo et al., 1997). The Sp3 factor binds to the GC box, and use of MGBs that preferentially bind this motif may also be useful in increasing gene expression and consequently enhancing chemosensitivity.

The initial report on topo II promoter sequences modulating down-regulation at confluence identified the ICB2 as the critical sequence modulating this effect (Isaacs et al., 1996). The MGB sequence-specificity is such that drug binding can be demonstrated over several ICBs and these experiments do not therefore rule out a contribution from other ICB sequences within the promoter. Cotransfection of a dominant negative NF-Y plasmid with a reporter construct containing a mutant ICB2 did not increase luciferase levels to the same level as that with constructs of the wild-type sequence. This could be because the transcription factor complexes with stimulatory effects are unable to bind to the promoter. Although there is data implicating ICB2 alone in transcriptional repression at confluence, there is evidence from other studies that several ICBs may be implicated in control of expression (Adachi et al., 2000). However, our data confirm that NF-Y is critical for transcriptional repression at confluence and prevention of its interaction with the promoter will alter gene regulation.

In conclusion, the topo II promoter is a valuable tool in investigation of the potential modulation of gene expression by inhibition of transcription factor binding. Strategies that modulate the activity of these transcription factors may also have a significant effect on gene expression.