Modulation of topoisomerase IIα expression by a DNA sequence-specific polyamide Free

Topoisomerase II (topo II) is an enzyme involved in several critical processes, including chromosome segregation, mitosis, and recombination (1). Two isoforms of topo II (α and β) have been described with differing biological functions, chromosomal locations, and patterns of expression. Expression of topo IIα varies within the cell cycle with increased expression at late S phase (2). Regulation of topo IIα expression occurs at transcriptional and posttranscriptional levels. Transcriptional down-regulation has been shown in conditions, including heat shock (3), p53 expression (4), and confluence (5).

Several chemotherapeutic drugs interact with topo II, including etoposide and doxorubicin. The mechanism of action of these agents is complex but there is interference with the religation step following formation of the cleavable complex resulting in cytotoxic DNA double-strand breaks (6). Increased cellular topo II expression results in sensitization to etoposide and doxorubicin (7, 8). Resistance to topo II poisons is associated with reduced transcriptional expression in vitro. For example, an etoposide-resistant cell line showed increased SP3 expression with repression of transcription (9). Decreased topo II expression in confluent cells contributes to cellular resistance to these agents (10, 11). Therefore, strategies that increase topo II transcriptional expression may sensitize cells to chemotherapy.

The promoter of the topo IIα gene has been well characterized and includes five CCAAT boxes in inverted orientation (ICB; Fig. 1A; ref. 12). Repression of the topo IIα gene under conditions of confluence is mediated through NF-Y interactions with ICBs within the promoter (13). The response of NIH3T3 cells was found to be similar to human cancer cells with arrest at confluence mediated through transcriptional inhibition; the murine promoter contains five ICBs (14).

The potential of small DNA-interactive molecules to modulate protein/DNA interactions within the topo IIα promoter and thereby chemosensitize confluent cells is being investigated. Several DNA minor groove binding drugs, including the oligopeptide antibiotic distamycin and the bis-benzimidazoles Hoechst 33342 and Hoechst 33258, have been shown to bind preferentially to AT-rich sequences in DNA (15). We previously investigated the potential to modulate topo II expression in confluent cells using Hoechst 33342 and Hoechst 33357 (16). These drugs blocked NF-Y/DNA interactions in an electrophoretic mobility shift assay (EMSA) and could displace bound transcription factors as determined by DNase I footprinting. When confluent cells with low levels of topo II were exposed to Hoechst 33342, there was a marked elevation of topo IIα expression associated with activation of the promoter. This resulted in increased sensitivity to etoposide. These experiments suggested the potential for increasing sensitivity to topo II poisons by modulating NF-Y binding to the topo IIα promoter. However, the toxicity of the Hoechst compounds and their lack of specificity reduce potential for further development.

There has been extensive interest in the synthesis of pyrrole and imidazole-containing polyamides to target specific DNA sequences (17). A hairpin polyamide molecule (JH-37) designed to recognize sequences encompassing ICB2/ICB3 within the promoter of the topo IIα gene was synthesized (18). Surface plasmon resonance studies confirmed that JH-37 binds more strongly to ICB2 than to ICB1. Thermal melting studies using synthetic DNA hairpins indicated larger ΔT_M values for ICB2 and ICB3 over ICB1, ICB4, and ICB5, indicating preferential binding to ICB2 and ICB3.

In this study, we show that JH-37 binds to the topo IIα promoter and inhibits NF-Y interactions with its cognate binding site. In confluent NIH3T3 cells incubated with JH-37, this results in up-regulation of topo IIα mRNA and protein expression with subsequent increase in etoposide-induced DNA double-strand breaks. Polyamides directed against the promoter of topo IIα may allow deregulation of gene expression and consequent increased chemosensitivity. These experiments also show the feasibility of designing polyamides to bind to short sequences within promoters, which could be used to modulate transcription.

JH-37 was synthesized as described previously (18) and dissolved in DMSO to a final concentration of 10 mmol/L. Etoposide was obtained from Sigma-Aldrich (Poole, Dorset, United Kingdom) and prepared in DMSO at a concentration of 100 mmol/L.

NIH3T3 cells (obtained from Cancer Research UK London Research Institute, London, United Kingdom) were grown in DMEM (Autogen Bioclear, Wiltshire, United Kingdom) supplemented with 10% newborn calf serum and 1% glutamine and incubated at 37°C in 5% CO₂. For studies on confluence, NIH3T3 cells were seeded at 5 × 10⁵ per flask in 80 cm² flasks and maintained in complete medium until they completely covered the flask. They were maintained at confluence for 96 h to allow topo IIα levels to decrease in the cells before releasing from confluence. JH-37 was then used to treat confluent cells, and samples were collected at 4, 6, and 24 h after release from confluence.

Nuclear extracts were prepared as described (19), and all steps were done at 4°C in the presence of a protease inhibitor mix (Complete, Roche Diagnostics GmbH, Mannheim, Germany).

The oligonucleotides (MWG Biotech, Ebersberg, Germany) containing ICBs (italicized) used in EMSAs are the following: topo IIα ICB1, 5′-CGAGTCAGGGATTGGCTGGTCTGCTTC-3′ (sense) and 5′-GAAGCAGACCAGCCAATCCCTGACTCG-3′ (antisense); ICB2, 5′-GGCAAGCTACGATTGGTTCTTCTGGACG-3′ (sense) and 5′-CGTCCAGAAGAACCAATCGTAGCTTGCC-3′ (antisense); ICB3, 5′-CTCCCTAACCTGATTGGTTTATTCAAAC-3′ (sense) and 5′-GTTTGAATAAACCAATCAGGTTAGGGGAG-3′ (antisense); and ICB4, 5′-GAGCCCTTCTCATTGGCCAGATTCCCTG-3′ (sense) and 5′-CAGGGAATCTGGCCAATGAGAAGGGCTC-3′ (antisense). 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 annealed in an equimolar ratio. Double-stranded oligonucleotides were 5′-end labeled with T4 kinase (Invitrogen, Paisley, United Kingdom) using [γ-³²P]ATP and subsequently purified on Bio-Gel P-6 columns (Bio-Rad, Hemel Hempstead, United Kingdom). EMSAs were essentially done as described (18). Briefly, ∼0.1 ng of radiolabeled probe was incubated for 2 h in a buffer containing 20 mmol/L K-HEPES (pH 7.9), 1 mmol/L MgCl₂, 0.5 mmol/L K-EDTA, 10% glycerol, 50 mmol/L KCl, 0.5 mmol/L DTT, and 0.5 μg poly(dI-dC).poly(dI-dC) (Pfizer, Tadworth, United Kingdom) and 1× protease inhibitor mix (Complete) at 30°C with the competitor JH-37 in this case. Crude cell lysate (20 μg) in a total volume of 8 μL was then added, and the reaction was further incubated at 30°C for 2 h. Subsequently, 0.5 μL of loading buffer [25 mmol/L Tris-Cl (pH 7.5), 0.02% bromphenol blue, 10% glycerol] was added and the samples were separated on a 4% polyacrylamide gel in 0.5× Tris-borate EDTA 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 (Kodak, Hemel Hempstead, United Kingdom).

A radiolabeled probe of 479 bp corresponding to positions −489 to −10 relative to the transcriptional start site of the topo IIα promoter was generated as follows: 4 pmol of the antisense oligonucleotide (5′-GTCGGTTAGGAGAGCTCCACTTG-3′) were 5′-end labeled with T4 kinase using [γ-³²P]ATP in a 10 μL reaction followed by heat inactivation for 20 min at 65°C. Subsequently, 4 pmol sense oligonucleotide (5′-CTGTCCAGAAAGCCGGCACTCAG-3′), 2 μL of 10 mmol/L deoxynucleotide triphosphate (Promega, Southampton, United Kingdom), 1 unit of Red Hot DNA polymerase (Abgene, Epsom, United Kingdom), 2 μL of 25 mmol/L MgCl₂, and 4.5 μL of 10× reaction buffer IV (Abgene) were added (in a final volume of 50 μL), and a PCR was done 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. DNase I footprint reactions were done with 30 μg nuclear extract in a 50 μL reaction in the same buffer as used for EMSA. After preincubation for 30 min at 4°C, ∼0.1 ng of radiolabeled probe was added and the mixture was incubated at room temperature for another 30 min. Subsequently, 1 unit of RQ1 DNase I (Promega) and up to 5 mmol/L of MgCl₂ and CaCl₂ were added. Following exactly 3 min of digestion at room temperature, 1 volume of stop mix containing 30 mmol/L K-EDTA (pH 8.0), 200 mmol/L 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 mmol/L K-EDTA (pH 8.0), 0.05% bromphenol blue, 0.05% xylene cyanol]. The sample was heat denatured for 3 min at 95°C and separated on a 6% denaturing polyacrylamide gel (SequaGel, National Diagnostics, East Riding, Yorkshire, United Kingdom). A 10-bp ladder (Invitrogen) labeled with ³²P by T4 kinase was used as a molecular weight standard. The dried gels were exposed to Kodak X-Omat-LS film with intensifying screens at −80°C.

For Western blot analysis, 50 μg of protein extract were denatured by heating for 3 min at 95°C in sample buffer containing 100 mmol/L Tris-Cl (pH 6.8), 4% SDS, 10% 2-mercaptoethanol, 20% glycerol, and 0.02% bromphenol blue. Bio-Rad high-range SDS-PAGE molecular weight 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, Sigma-Aldrich). Western blot analysis was done with the IHIC8 rabbit polyclonal topo IIα antibody (kindly provided by Dr. I.D. Hickson, Weatherall Institute of Molecular Medicine, Oxford, United Kingdom) at a 1:5,000 dilution using an enhanced chemiluminescence Western blot detection kit and protocol (GE Healthcare UK Ltd., Buckinghamshire, United Kingdom) using 1% skimmed milk as blocking reagents and TBS plus 0.5% Tween 20 (Sigma-Aldrich) as a buffer. The chemiluminescent signal was visualized by exposing the blots to Kodak X-Omat-LS film.

Reverse transcriptase was carried out essentially as described in the manufacturer's instructions. Briefly, RNA was extracted from cells using the RNeasy Mini kit (Qiagen, Crawley, West Sussex, United Kingdom). Samples were resuspended in RLT buffer before homogenizing and applying to the supplied columns. The bound RNA was washed with buffer RPE and eluted in nuclease-free water. The concentration of purified RNA was determined by measuring the absorbance at 260 nm. Subsequently, the reverse transcription reaction was carried out at 48°C for 45 min using 5 μg RNA, 4 μL of avian myeloblastosis virus reverse transcriptase enzyme (Promega), 2 μL RNasin RNase inhibitor (Promega), 8 μL reverse transcriptase buffer, 4 μL of 10 mmol/L deoxynucleotide triphosphate, 8 μL oligo(dT)_12-18 (Invitrogen), and nuclease-free water in a final volume of 40 μL. Avian myeloblastosis virus reverse transcriptase enzyme was inactivated by heating the reaction mix at 94°C for 2 min.

Real-time PCR was carried out using the ABI PRISM 7000 Sequence Detection System from Applied Biosystems (Warrington, United Kingdom). Respectively, the forward and reverse topo IIα primers used were 5′-GATTCATTGAAGACGCTTCGTTAT-3′ and 5′-GATGGATAAAATTAATCAGCAAGCCT-3′. The probe sequence (CAGATCAGGACCAAGATGGTTCCCACATC) used for the reactions was labeled at the 5′-end with 6-FAM and TAMRA at the 3′-end. The cycling conditions used were 50°C for 2 min and 95°C for 10 min to allow denaturation to occur and 40 cycles of 95°C for 15 s and 58°C for 1 min to amplify the target sequences. Glyceraldehyde-3-phosphate dehydrogenase primer/probe master mix (1.25 μL; Applied Biosystems) was used as an internal control in all reactions. The reaction mix was prepared using 12 μL of the Taqman PCR master mix (Applied Biosystems) and 1 μmol/L of each primer, 0.2 μmol/L probe, and 2.5 μL of cDNA template in a final volume of 25 μL. The results were analyzed using the mathematical quantification approach described by Pfaffl (20) and ABI User Bulletins nos. 2 and 5 (2001; ref. 17). This is based on the relative expression ratio of the target gene (topo IIα) compared with that of an internal control gene (glyceraldehyde-3-phosphate dehydrogenase). Standard curves were constructed for both the internal and reference genes, and slopes of these were used to ensure that both primer sets were equally efficient. The threshold cycle values (C_t) and the efficiencies of the reactions were used to compare the relative expression levels of the target gene in various samples. To ease comparison, levels of topo IIα RNA in untreated, exponentially growing cells were set at a value of 1 and all test samples were expressed at values relative to this. Statistical analysis was carried out using the Student's t test.

Immunoprecipitations were carried out essentially as described with modifications (21). Cells were cultured and treated in 150-mm plates and treated with 1% formaldehyde to induce the cross-linking reaction. Treatment with 0.125 mol/L glycine stopped the reaction, and cell pellets were stored at −20°C until analysis. To analyze, cells were resuspended in lysis buffer [5 mmol/L PIPES (pH 8.0), 85 mmol/L KCl, 0.5% NP40, 1× protease inhibitor cocktail (Sigma-Aldrich)] containing 0.5 mmol/L phenylmethylsulfonyl fluoride. Subsequently, nuclei extracted using a Dounce homogenizer were resuspended in sonication buffer [50 mmol/L Tris-HCl (pH 8.0), 10 mmol/L EDTA, 0.1% SDS, 0.5% deoxycholic acid, 1× protease inhibitor cocktail] and sonicated into 500- to 1,500-bp chromatin fragments. The chromatin fragments were stored at −80°C pending further analysis. Protein G (15 μL; Kierkegaard and Perry Lab., Inc., Gaithersburg, MD) was precleared overnight with 1 μg/μL salmon testis DNA and 1 μg/μL bovine serum albumin in immunoprecipitation buffer [50 mmol/L Tris-HCl (pH 8.0), 10 mmol/L EDTA, 0.1% SDS, 0.5% deoxycholic acid, 1× protease inhibitor cocktail, 150 mmol/L LiCl]. Chromatin (25–50 μL) was also precleared by incubating for 2 h with 40 μL of protein G slurry in immunoprecipitation buffer at 4°C. The precleared chromatin was placed in presiliconated 0.5 mL PCR tubes, up to 8 μg of antibody were added (200 μL final volume), and the mixture was incubated overnight at 4°C. Subsequently, 110 μL of the salmon testis DNA–saturated and bovine serum albumin–saturated protein G in immunoprecipitation buffer were added to the chromatin-antibody mixture, and the samples were further incubated for 2 h at 4°C. The samples were centrifuged at 4,000 rpm for 2 min, and the supernatant was stored at −20°C as a source of ‘input DNA’. The resin was washed initially at 4°C for 30 min using 300 μL of immunoprecipitation buffer. Subsequently, nine more washes were carried out by resuspending the resin in 300 μL of immunoprecipitation buffer and centrifuging for 2 min at 4,000 rpm. The bound DNA was then eluted from the resin by adding 100 μL of elution buffer (1% SDS, 50 mmol/L NaHCO₃, 1.5 ng/μL salmon testis DNA) and incubating for 1 h at 37°C on a shaker. After centrifugation at 14,000 rpm for 2 min, the supernatant and the input DNA were both incubated overnight at 65°C with 10 μg RNase A and 200 mmol/L NaCl to reverse the cross-links. Following this, the DNA was precipitated with 99% ethanol at −20°C. The pellets were collected by centrifugation at 13,000 rpm for 30 min, washed with 70% ethanol, and air dried. The protein was removed from the DNA by resuspending the pellets in 40 μg proteinase K, 25 μL of proteinase K buffer [1.25% SDS, 50 mmol/L Tris (pH 7.5), 25 mmol/L EDTA], and 100 μL Tris-EDTA (pH 7.5) and incubating at 42°C for 2 h. Digested protein was removed with phenol-chloroform-isoamyl alcohol (25:24:1), and the DNA was precipitated at −20°C overnight with 30 μL of 3 mol/L sodium acetate, 1 μL of 5 mg/mL tRNA, and 750 μL of 99% ethanol. The sample DNA pellets were resuspended in 60 μL sterile water and the input DNA in 200 μL. The DNA was then used for PCR using 2 μL DNA/sample.

The comet assay was carried out essentially as described (22). NIH3T3 cells, maintained at confluence for 96 h, were treated with either etoposide (1, 10, or 100 μmol/L), JH-37 (1 or 10 μmol/L), or a combination of the two compounds, where each concentration of JH-37 was combined with each of the concentrations of etoposide. Cells were treated with etoposide alone for 2 h, JH-37 alone for 4 h, and, for the combination treatment, initially with JH-37 for 4 h followed by a simultaneous treatment with JH-37 and etoposide for 2 h. The treated samples were stored at −80°C until analysis. Washed slides were dried at 37°C, and comets were visualized and measured using the Komet 4 comet imaging system (Kinetic Imaging, Nottingham, United Kingdom) with fluorescent inverted microscope (Nikon, Surrey, United Kingdom).

Confluent cells were treated with 100 μmol/L etoposide and 10 μmol/L JH-37 in combination using the same treatment schedule as described above. Cells were then maintained in medium containing 10 μmol/L JH-37, and at 6 and 24 h following treatment, cells were stained with trypan blue solution (Sigma-Aldrich) and counted to determine the viability of cells following treatment with etoposide, JH-37, or a combination of both.

The structure of JH-37 and its chemical properties have been described elsewhere (Fig. 1B; ref. 18). JH-37 is designed to target the sequence 5′-TTGGT-3′ encompassing ICB sequences of the topo IIα promoter (Fig. 1A). The topo IIα promoter consists of five ICBs of which the proximal three are necessary as a minimal promoter (12).

Our previous studies indicated that, although NF-Y has been shown to stimulate the topo IIα promoter under several conditions, including cell cycle progression, it inhibits promoter activity in confluent NIH3T3 cells (16). To determine if JH-37 could inhibit binding of transcription factors to ICB motifs within the topo IIα promoter, EMSA studies were carried out using radiolabeled probes containing nucleotide sequences of ICB1, ICB2, ICB3, and ICB4 incubated with nuclear extracts from confluent NIH3T3 cells. As shown in Fig. 2, inhibition of protein binding was shown at 1 to 5 μmol/L for ICB1, ICB2, and ICB3 and between 10 and 50 μmol/L for ICB4. No effect on band intensity was found when reactions were incubated with an excess of unlabeled mutated oligonucleotide, although bands were competed out with an excess of unlabeled wild-type oligonucleotides showing specificity of binding. The complex binding ICBs contained NF-Y as shown by supershifting with antibody to the A subunit of NF-Y.

To identify the pattern of JH-37 binding to specific sequences in the topo IIα promoter, in vitro DNase I footprinting was carried out using a radiolabeled probe of the topo IIα promoter incubated with drug. With increasing concentration of JH-37, there was clear protection over the ICB1, ICB2, and ICB3 sequences (Fig. 3A). With 0.5 μmol/L JH-37, there was protection over ICB2 and ICB3 sites, whereas 1 μmol/L showed protection over the ICB1 site. Concentrations at ≥5 μmol/L showed larger footprints, suggesting more than one binding site.

Although several proteins, including cut and YB-1, have been identified as binding to CCAAT motifs, the NF-Y complex has been most extensively characterized. To investigate if the displacement of the complex containing CCAAT-binding proteins by minor groove binding drugs could be shown within the topo IIα promoter, in vitro DNase I footprinting was carried out with recombinant NF-Y protein. A radiolabeled probe containing the topo IIα promoter was incubated with NF-Y protein in the presence of increasing amounts of drug (Fig. 3B). The pattern of protection over the three ICBs was altered with increasing concentration of JH-37, resulting in footprints identical to that found with JH-37 and DNA alone. This indicates that incubation with JH-37 can displace the CCAAT-binding proteins. This effect was seen when incubation with JH-37 occurred before or after addition of NF-Y protein (data not shown).

A potential problem with polyamides is lack of nuclear penetration. Fluorescent-labeled conjugates have been used to investigate nuclear permeability (23). However, the chromatin immunoprecipitation (ChIP) assay allows quantitative assessment of promoter occupancy in intact cells. Therefore, to determine if JH-37 could bind to NF-Y sites in vivo and inhibit binding of NF-Y complexes to the topo IIα promoter, ChIP assays were carried out using antibodies to the NF-YB subunit of the NF-Y complex as well as with control anti-FLAG antibodies. NIH3T3 cells were used in these experiments because of the reproducibility of confluence-induced arrest and down-regulation of topo IIα expression, which is comparable with human cells (14). The promoter region has a high degree of homology with the human promoter (14).

Chromatin from exponential phase cells showed a modest enrichment of NF-Y over control antibody; confluent NIH3T3 cells showed increased NF-Y binding to the topo IIα promoter (Fig. 4). Thus, the nonspecific FLAG band can be visualized and quantitated at 33 cycles in extracts from confluent cells at levels that are comparable with YB expression levels at 28 cycles, which indicates >30-fold enrichment. The increased NF-Y association with the topo IIα promoter in confluent cells is consistent with a specific negative role of the trimer on the promoter. In Fig. 4A, representative relative enrichments with anti-YB are compared with the negative anti-FLAG control on three different promoters. Binding to the topo IIα promoter showed a dramatic change under the cellular conditions analyzed. On treatment of confluent NIH3T3 cells with JH-37, there was a significant reduction in NF-Y binding to the topo IIα promoter. There is minimal binding of NF-Y to the cdc2 promoter in confluent cells, which is not significantly altered following exposure to JH-37. ChIPs were normalized with the fos target, which lacks NF-Y-binding sites, to score for global recovery of DNAs. Quantification of this analysis is shown in Fig. 4B. Following exposure of confluent NIH3T3 cells to 10 μmol/L JH-37, there was a 6-fold reduction in NF-Y binding to the promoter. Thus, JH-37 has clear inhibitory effects on NF-Y binding to the topo IIα promoter.

The effects of JH-37 on topo IIα expression were analyzed in confluent NIH3T3 cells, in which there is transcriptional down-regulation of topo IIα mRNA and protein. Following incubation of confluent NIH3T3 cells with 10 μmol/L JH-37, there was a 6-fold increase in topo IIα mRNA at 6 h rising to 10-fold at 24 h compared with untreated confluent cells (Fig. 5). Statistical analysis yielded a P value of 0.01. This was associated with increased topo IIα protein expression detectable at 6 h. Experiments done with the human colon cancer cell line CaCo2, which also shows decreased topo IIα transcriptional expression at confluence, showed similarly increased mRNA and protein expression following exposure to JH-37 (data not shown).

Decreased topo IIα expression in confluent cells results in resistance to the topo II poison etoposide, and increased expression of topo IIα can sensitize cells. Having confirmed that increased topo IIα mRNA and protein expression could be shown in confluent cells following drug treatment, the effects of JH-37 treatment on protein-associated DNA strand breaks were investigated. Comet assays were done following a 2-h exposure of cells to etoposide at various concentrations (Fig. 6A). No DNA strand breaks were detectable in cells treated with JH-37 alone. A dose-dependent increase in DNA strand breaks was detectable in confluent cells treated with etoposide. However, in cells incubated with a combination of 10 μmol/L JH-37 and etoposide, a 56% increase in the tail moment was detectable by the comet assay. This shows that increased topo IIα expression following JH-37 treatment is associated with an increased formation of DNA strand breaks following exposure to etoposide.

We further investigated the effects of combining JH-37 and etoposide treatments by assessing the viability of cells after treatment (Fig. 6B). These cells are highly resistant to etoposide treatment; in contrast, exposure of exponentially growing cells to etoposide results in high levels of DNA strand breaks (data not shown). Treatment of confluent phase NIH3T3 cells with 10 μmol/L JH-37 alone caused a 2 ± 1% reduction in cell viability, whereas 100 μmol/L etoposide alone caused a 6 ± 1% decrease in viability at 24 h. Combined treatment with 10 μmol/L JH-37 and 100 μmol/L etoposide resulted in 30.0 ± 2% loss of viability compared with the single agent treatments at 24 h. This increased to 51 ± 1% and 75 ± 1% at 48 and 72 h, respectively. In contrast, exposure to etoposide alone resulted in 28 ± 3% and 44 ± 1% decrease in viability at these time points. This indicates that combined treatment sensitizes cells in a synergistic manner, causing greater reduction in cell viability than when either compound is used alone.

This study shows that sequence-specific polyamides can be used to block transcription factor/DNA interactions and thereby modulate sensitivity of cells to chemotherapeutic agents. Although it has been estimated that it might be necessary to bind between 16 and 18 nucleotides to achieve complete gene specificity, it is clear that such specificity is not required to show biological effects in vivo.

The topo IIα promoter is among several TATA-less promoters that are regulated by the NF-Y complex. The NF-Y complex consists of three subunits (NF-YA, NF-YB, and NF-YC; ref. 24). Dynamic regulation of NF-Y binding to specific promoters within different phases of the cell cycle has been shown by ChIP studies (21). Transcriptional regulation of topo IIα in response to cell cycle progression (25, 26), effects of p53 (4, 27), and heat shock (3) is mediated through inverted CCAAT sequences within the promoter. Although the effects of NF-Y are stimulatory, there is evidence that negative regulation may occur following NF-Y interaction with the promoter. For example, the inhibitory effects of p53 on a variety of G₂-M promoters are mediated via NF-Y sites (28). Additionally, NF-Y binding to the ICB1 sequence within the topo IIα promoter is repressive in relation to cell cycle–associated transcription.

In this study, we investigated the potential for a polyamide, which blocks NF-Y/DNA interactions, to up-regulate topo II expression in confluent cells. The involvement of NF-Y in transcriptional down-regulation of topo IIα in confluent cells was shown in EMSAs and by transfection of a dominant-negative NF-Y, which resulted in increased expression of topo IIα promoter reporter constructs (16). Our results indicate a good correlation between inhibition of binding in vitro and in vivo and up-regulation of topo IIα. It is important to note the 10-fold increase of NF-Y binding under conditions of confluence. The ChIP assay was used to assess inhibition of NF-Y binding to the topo IIα promoter; current experiments are focussing on investigating the effects on individual ICBs.

Previously, we showed that the bis-benzimidazole compounds Hoechst 33342 could abrogate NF-Y binding to the topo IIα promoter, thereby increasing topo IIα expression and increasing sensitivity to the topo II poison etoposide (16). However, these agents are nonspecific in DNA-binding sequence requirements and, additionally, have multiple intracellular targets, including topoisomerase I (29, 30). Other compounds have been shown to block NF-Y binding, including HMN-176 [(E)-4-{[2-N-[4-methoxybenzenesulfonyl]amino]-stilbazole}1-oxide], an active metabolite of the synthetic antitumor compound HMN-214 [(E)-4-{2-[2-(N-acetyl-N-[4-methoxybenzenesulfonyl]amino)stilbazole]}1-oxide], which inhibits multidrug resistance 1 expression by its effect on blocking NF-Y binding to CCAAT sequences within the promoter (31).

The concept of sequence specificity achieved by polyamides is based on the pioneering work of Dervan et al. (32, 33). The ability of these compounds to block DNA interactions in some genes at nanomolar concentrations has been shown in vitro, although higher concentrations have been used in vivo. However, there have been problems in showing that these compounds are able to penetrate the nucleus and act on their predicted sequences. Clearly, some polyamides are able to enter the nuclei of cells in vitro as determined by fluorescein-labeled conjugates (23). This is an increasing problem as the size of the polyamide is increased. Accessibility of nuclear chromatin was shown by a polyamide-chlorambucil conjugate, which generated distinct transcription profiles at concentrations of 2 μmol/L (34). In view of the difficulty of nuclear penetration for polyamides targeting larger number of nucleotides, it may be necessary to use these agents in combination.

The study reported here shows that a polyamide designed to target five nucleotides can bind to the topo IIα promoter and modulate transcriptional expression. Although JH-37 binds preferentially to the 5′-TTGGT-3′ sequence that overlaps ICB2 and ICB3, it is not surprising that it also binds ICB1, which overlaps on a 5′-TTGGC-3′ site. It has been shown that the γ-aminobutyrate linker unit exhibited minimal preference for an A.T base pair over a G.C base pair (32). In the system used in this study, increased topo IIα expression is associated with increased number of DNA strand breaks and thereby increased chemosensitivity. Although many factors influence cellular sensitivity to etoposide, there are clear correlations with levels of topo II expression in vitro (7, 8). The relatively modest decrease in cell viability found in our study, in contrast to the effects on gene expression, is due to the lack of cycling cells in the confluent population, as the cytotoxic effects of topo II poisons are influenced by replication fork collision (35). Although design of a polyamide targeted toward increased numbers of nucleotides would improve specificity, this could compromise the nuclear penetration achieved with JH-37. A recent study using a pyrrole-imidazole polyamide directed against the rat transforming growth factor-β1 promoter showed nuclear penetration and inhibition of gene expression following i.v. injection (36, 37). An alternative strategy is the use of multiple polyamides with overlapping nucleotide preference over the promoter, which we are currently investigating.