Introduction

Top Summary Introduction Materials & Methods Results Discussion References

DNA topo II (EC 5.99.1.3), a nuclear enzyme that controls DNA topology, is the target for many anti-cancer drugs. The enzyme exists as two distinct isoforms (termed  and ) that differ in many respects, including cell cycle regulation and nuclear localization (Wang, 1996; Austin and Marsh, 1998). The enzyme modifies DNA structure by making a double strand break, thus creating a "gate" through which another DNA duplex can pass. After strand passage, the break is ligated and the DNA structure is restored (Liu et al., 1980). This process facilitates cellular processes that require modifications in DNA topology, including replication, transcription, and repair. Many topo II-targeting anti-cancer agents are cytotoxic via stabilization of the normally transient "cleavable" complex. In the drug-stabilized cleavable complex, topo II molecules are covalently bound to DNA strand ends at the double strand break, thereby preventing religation of the DNA strands (Osheroff, 1989). These stabilized complexes are thought to be lethal upon DNA replication when the break becomes permanent, leading to mutations, recombination events, and chromosome aberrations (Chen et al., 1996; Suzuki et al., 1997).

Development of resistance to topo II-targeting drugs is a major clinical problem, and cell line studies have revealed that resistance may be dependent on many factors, including mutations in the topo II genes (Patel and Fisher, 1993; Dereuddre et al., 1995) resulting in decreased catalytic activity (Charcosset et al., 1988) or decreased protein levels (Mirski et al., 1993), increased drug efflux by overexpression of the mdr1 and MRP genes (Long et al., 1991; Lorico et al., 1995), and/or alterations in DNA repair rates (Ritke et al., 1994).

One approach to studying topo II drug sensitivity in vivo has been the use of a yeast model system, in which yeast carrying a topo II temperature-sensitive mutation can be rescued by plasmid-borne cDNAs encoding human topo II or  (Meczes et al., 1997). This system highlighted differences in drug sensitivities between the two isoforms and will be useful for analysis of drug resistance mutations in topo II genes and screening for isoform-specific drugs. Another approach is the in vitro cleavage assay, which has been used to analyze the sequence specificity of both isoforms in the presence of anti-topo II agents. Using this assay it has been shown that both the  and  isoforms formed cleavable complexes in vitro with mAMSA, teniposide and anthracyclines (Cornarotti et al., 1996; Marsh et al., 1996).

The most well documented model is that of mammalian cell lines, and various techniques have been used to estimate cellular topo content and activity in cell lines and clinical samples. These methods include preparation of whole cell or nuclear extracts for immunoblotting and decatenation/relaxation assays (Danks et al., 1988; Drake et al., 1989) and measurement of topo II mRNA and protein levels (Brown et al., 1995; Houlbrook et al., 1995). These data give an indication of topo abundance in vitro, but protein levels do not necessarily reflect enzyme activity or ability to interact with the drug in vivo. The extent and type of DNA strand breaks after treatment with anti-topo II agents are often measured by alkaline elution (Pommier et al., 1994). This technique has been used to show that etoposide introduces protein-associated strand breaks in a dose-dependent manner, reaching a maximum after 2 hr of treatment (Kalwinsky et al., 1983; Long et al., 1985). Alkaline elution is very sensitive and has proved invaluable in the analysis of strand breaks; however, it does require a relatively large number of cells, and it cannot distinguish between breaks mediated via DNA topo II or .

We report the development of a new in situ assay to quantify drug-stabilized cleavable complex formation in individual cells by immunofluorescence with isoform-specific antibodies. The assay was adapted from the method for quantification of melphalan-DNA adducts (Frank et al., 1996). We have focused on the effects of etoposide (VP-16, a widely used epipodophyllotoxin) on human leukemic CCRF-CEM cells. This technique has a number of advantages. It detects DNA topo II covalently linked to DNA in drug-stabilized, cleavable complexes (rather than strand breaks revealed after proteinase K treatment); the use of isoform-specific antisera allows cleavable complexes containing either DNA topo II or to be distinguished; and it requires only small numbers of cells.

	Materials and Methods

Top Summary Introduction Materials & Methods Results Discussion References

Cell culture. The CCRF-CEM human leukemic cell line was maintained as a suspension culture at 37° (in a humidified atmosphere of 5% CO₂) in RPMI 1640 medium supplemented with 10% fetal bovine serum and penicillin (50 units/ml)/streptomycin (50 µg/ml) (cell culture reagents from Gibco/BRL, Paisley, UK). The cells were found to be free from Mycoplasma sp. contamination.

Drug treatments and irradiation. Etoposide (1 mg/ml stock dissolved in methanol and stored at 20°) was purchased from Sigma (Poole, Dorset, UK). Exponentially growing cells (~4 × 10⁵/ml) were exposed to etoposide for 2 hr. They were then washed by centrifugation (1000 × g, 5 min) twice and resuspended in PBS (2.7 mM KCl, 13.7 mM NaCl, 1.5 mM KH₂ PO₄, 8 mM Na₂ HPO₄, pH 7.2, 0°). To study loss of cleavable complexes, drug-treated cells were washed by centrifugation and incubated in fresh medium for various time periods before being finally washed and resuspended in PBS (0°). For irradiation experiments, cells were irradiated in PBS (0°) at 4 Gy/min using a ¹³⁷Cs source (Gammacell 1000; Nordion Int., Kanata, Ontario, Canada).

Antibodies. Anti-topo II polyclonal antibodies raised in rabbits were used for these experiments. 18511 was raised to recombinant human topo II and 18513 to a recombinant human topo II carboxyl-terminal fragment. Western blots demonstrated that 18511 detected the isoform specifically and 18513 detected the  isoform specifically (data not shown). In these assays, 18511() was used at a 1:50 dilution and 18513 () at 1:250. The anti-rabbit FITC-conjugated second antibody (F(ab')₂ fragment (Sigma, Poole, Dorset, UK) was used at 1:100 dilution. The mouse anti-histone antibody, which detects H1, H2A, H2B, H3, and H4 (Boehringer Mannheim, Germany), was used at a 1:30 dilution, and the anti-mouse FITC conjugate (Sigma, Poole, Dorset, UK) was used at 1:100.

Agarose embedding and staining. The method for precoating microscope slides and embedding cells in agarose was as described previously (Frank et al., 1996). In brief, 50 µl of cell suspension was warmed to 37° and immediately mixed with an equal volume of agarose solution [2% (w/v) in PBS; "SeaPrep" ultra-low gelling; FMC Bioproducts, Rockland, ME], which had been melted and kept at 37°. The mixture was immediately spread evenly across a slide, and the agarose was quickly gelled by placing the slides on a cold surface (0°). The slides were then immersed in lysis buffer for 15 min at 20°. This lysis buffer contained 1% (w/v) sodium dodecyl sulfate, 80 mM phosphate buffer, pH 6.8, 10 mM EDTA, and a protease inhibitor mixture (2 µg/ml pepstatin A, 2 µg/ml leupeptin, 1 mM phenylmethylsulfonyl fluoride, 1 mM benzamidine, and 1 mM dithiothreitol, final concentrations). In the lysis buffer, 1% Sarkosyl may be successfully substituted for 1% sodium dodecyl sulfate. At this stage slides could be stored at 20° in PBS containing 10% glycerol without loss of signal. Slides were next immersed in 1 M NaCl supplemented with the protease inhibitor mixture for 30 min at 20°, and then washed by immersion three times (5 min per wash) in PBS. Slides were exposed to primary antibody (diluted in PBS containing 0.1% (v/v) Tween 20 and 1% (w/v) bovine serum albumin) for 1 hr in a humidified atmosphere, then washed three times in PBS containing 0.1% (v/v) Tween 20 (PBST). Slides were similarly exposed to the FITC-conjugated second antibody (diluted in PBS containing 0.1% (v/v) Tween 20 and 1% (w/v) bovine serum albumin) for 1 hr before two washes in PBST and one final wash (30 min or overnight at 4°) in PBS containing the protease inhibitor mixture. It was found that increasing the length of incubation time for primary and/or secondary antibodies did not increase fluorescence levels (data not shown), indicating saturation of antibody binding within 1 hr. Slides were stained with Hoescht 33258 (10 µM in PBS; Sigma, Poole, Dorset, UK) for 5 min before application of coverslips that were secured with a sealant. We have termed this technique TARDIS.

Quantitative fluorescence microscopy and image analysis. This was described in detail previously (Frank et al., 1996). In brief, an epifluorescence microscope (Olympus BH2-RFCA, 10× objective) was used to separately visualize the blue (hoechst-stained DNA) fluorescence and the green (FITC-stained topo II) immunofluorescence. Images were captured at an accuracy of 16 bits per pixel using a cooled slow-scan charge-coupled device camera (Astrocam, Cambridge, UK). DNA in a particular field of view was focused only under blue fluorescence to minimize photo bleaching of FITC. An image was then captured of blue fluorescence (5-sec exposure) and then a further image of the same field of view was captured of green fluorescence (20-sec exposure) by using specific filter sets. An average of eight pairs of images per drug dose were captured from replicate slides for each antibody (this gave a total of 100-150 cells per dose for each antibody)

Cytotoxicity assay. CCRF-CEM cells were seeded (5 × 10⁴/ml) into 6-well plates. After 48 hr, the cells were exposed to etoposide for 2 hr, then washed in PBS and incubated in fresh medium. After 5 days, cells were counted by trypan blue exclusion, and the results were expressed as described previously (Boulton et al., 1995) as a percentage of growth inhibition compared with controls.

	Results

Top Summary Introduction Materials & Methods Results Discussion References

Immunofluorescent staining of cleavable complexes via agarose embedding. CCRF-CEM cells were exposed to a range of concentrations of etoposide for 2 hr. The newly developed assay (illustrated in Fig. 1) was used to detect drug-stabilized topo II- and -cleavable complexes in individual cells from these cultures. Cells (with/without drug treatment) were embedded in agarose on microscope slides and lysed to remove the cell membrane and soluble proteins. Salt extraction removed nuclear proteins including noncovalently bound topo II. Drug-stabilized covalent topo II-DNA complexes remained and were detected by staining replicate slides (from the same culture) using either 18511() or 18513() antisera followed by an FITC-conjugated second antibody.

View larger version (31K): [in this window] [in a new window]   Fig. 1.   Schematic diagram of the agarose-embedding method to detect DNA topo II-cleavable complexes.

View larger version (49K): [in this window] [in a new window]   Fig. 2.   Fluorescence of etoposide-treated CCRF-CEM cells. Cells were treated with 0 (A, B), 1 (C, D) 10 (E, F), or 100 µM (G, H) etoposide for 2 hr before embedding and staining with 18511() Images A, C, E, and G show the Hoechst-stained DNA (blue), whereas images B, D, F, and H show the corresponding FITC-stained (green) immunofluorescence due to etoposide-stabilized cleavable complexes. (All images were captured under 10× objective and were typical of those seen in replicate experiments). Bar, 100 µM.

Quantification and analysis of results. Digital images of Hoechst (DNA) fluorescence and FITC immunofluorescence were subjected to correction and analysis procedures to quantify the blue and green fluorescence associated with the DNA from each cell. Detailed results for a typical experiment in which the DNA was stained with 18511() are shown in Figs. 3 and 4. Fig. 3, a and b, shows scattergrams in which every cell that was analyzed is depicted as a dot. This representation of results shows not only the wide range of fluorescence values for Hoechst (DNA) staining, but also the existence of two distinct populations of cells at the highest doses (Fig. 3a). FITC immunofluorescence values (Fig. 3b) did not vary considerably for the untreated cells and 0.1 µM dose, but did begin to show a wide spread of values for the 10 and 100 µM doses.

View larger version (22K): [in this window] [in a new window]   Fig. 3.   Dose-dependent nature of fluorescence after etoposide treatment. Cells were treated with 0, 0.1, 1, 10, and 100 µM etoposide for 2 hr, and then slides were prepared. These were stained with 18511() and processed and analyzed as described in the text. Scattergrams show fluorescence levels for each individual cell for Hoechst (a) or FITC (b). Plots of median fluorescence values are shown for Hoechst (c) or (d) FITC, with the drug dose shown on a log scale for clarity. Integrated fluorescence values on all y-axes have been reduced for simplicity by 104. The data were obtained from one experiment typical of five other independent experiments.

View larger version (10K): [in this window] [in a new window]   Fig. 4.   Histograms show the frequency distribution of immunofluorescence for untreated cells (a) or for cells treated with 1 µM (b) or 100 µM (c) etoposide and stained with 18511(). The data were obtained from one experiment which was typical of five others.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Summary of data of three to six independent experiments. Cells were treated exactly as described in Fig. 3, but here data are presented for both antibodies. Plots show means ± standard error bars calculated from median fluorescence values obtained from independent experiments. a and c, Hoechst fluorescence for cells treated with 18511() or with 18513() respectively; b and d, FITC immunofluorescence values for 18511() or 18513(), respectively.

Specificity of complex recognition. To confirm that the immunofluorescence detected by this technique was specific for topo II cleavable complexes, the following experiments were performed.

View larger version (19K): [in this window] [in a new window]   Fig. 6.   Disassociation of topo II-cleavable complexes with time. a, Cells were treated with etoposide (0, 10, and 100 µM) for 2 hr, then slides were prepared immediately () or after drug was washed out and cells incubated in fresh medium for 30 (), 60 (), or 120 () min. Slides were stained with 18511() and processed as described in the text. b, Cells were similarly treated with etoposide, and slides were prepared immediately () or after incubation in drug-free medium for 15 (), 30 (), or 60 () min. Plot shows mean and mean ± standard error bars for immunofluorescence values obtained from four (a) or three (b) independent experiments.

Cytotoxicity. Fig. 7 shows the cytotoxic effect of etoposide on CCRF-CEM cells after a 2-hr exposure. The IC₅₀ value for etoposide is ~1 µM, and there were no surviving cells at either 10 or 100 µM.

View larger version (14K): [in this window] [in a new window]   Fig. 7.   CCRF-CEM cells were treated with 0, 0.1, 1, 10, or 100 µM etoposide for 2 hr. After exposure, the drug was washed off, and cells were incubated in fresh medium for 5 days. Results are expressed as described in the text. Plot shows the means and mean ± standard error bars obtained from four independent experiments.

	Discussion

Top Summary Introduction Materials & Methods Results Discussion References

We have shown for the first time that both isoforms of human DNA topo II are targeted by etoposide in vivo. This report describes a novel "TARDIS" method with which etoposide-induced topo II- and -cleavable complex formation in individual cells can be visualized and quantified. Using this method, we have shown that both isoforms form cleavable complexes in vivo after etoposide treatment in a dose dependent manner. This confirms the previously obtained in vitro data (Cornarotti et al., 1996; Marsh et al., 1996) and data from a yeast model system (Meczes et al., 1997), which indicated that both isoforms were potential targets for etoposide.

The results showed that the increase in immunofluorescence intensity was dependent on etoposide concentration (Figs. 3 and 4), becoming highly significant at 10 and 100 µM doses (p < 0.0001, Mann-Whitney test) The actual values for immunofluorescence (at 100 µM etoposide) were approximately 5-fold higher for the 18511() antibody than they were for the 18513() antibody (Fig. 5, b and d). This initially suggests that the  isoform may be a better target for etoposide in these cells, possibly because it is more sensitive to stabilized complex formation. However, it cannot be assumed that the higher values for the  antibody directly reflect actual enzyme quantities, because the efficacy of one antibody compared with the other is not known.

Several observations confirmed that the immunofluorescence measured in this assay was specific for topo II-cleavable complexes. First, it is well documented that the cleavable complexes are reversible. After removal of the drug, the equilibrium between topo II trapped in cleavable complexes and unbound topo II carrying out its normal cleavage/religation reaction shifts back toward unbound topo II (Hsiang and Liu, 1989). This manifests itself in the loss of drug-induced DNA strand breaks (Caldecott et al., 1993) and in this case, in a time-dependent decrease in immunofluorescence consistent with the disassociation of cleavable complexes.

Second, 18511() and 18513() antisera bound specifically to topo II proteins stabilized in the cleavable complex and not to other DNA strand breaks, as illustrated by the use of irradiated cells. Third, the staining of drug-treated cells with an anti-histone antibody did not result in any detectable levels of immunofluorescence, demonstrating that the lysis procedure removes histones and probably other nuclear proteins, leaving only stabilized topo II-cleavable complexes.

In addition to the significant increase in immunofluorescence after etoposide treatment, there was also a small increase in the Hoechst fluorescence (Fig. 5, a and c). However, this increase reached a plateau by 1 µM and was not dose-dependent at higher etoposide concentrations. Hoechst is a minor groove binding dye that binds most strongly to double-stranded DNA with a preference for AT-rich sequences (Vega et al., 1996 and references therein). The increase in Hoechst fluorescence was presumably caused by a change in DNA conformation resulting from inhibition of topo II, because the 2-hr etoposide exposure period was too short to permit significant accumulation of G₂ cells and concomitant increase in DNA content.

The IC₅₀ value for growth inhibition by etoposide was found to be 1 µM (Fig. 7). The results from Figs. 3 and 4 demonstrated that cleavable complexes became detectable at 1 µM, suggesting that relatively low levels of etoposide-induced complexes trigger events leading to cell death. This implies that the cleavable complex is a highly cytotoxic lesion, but it is likely that events downstream of complex formation [e.g., inhibition of replicon initiation (Kaufmann et al., 1991) DNA strand breaks, sister chromatid exchanges (Chaterjee et al., 1990), and other chromosome aberrations (Chen et al., 1996)] are ultimately responsible for cell death. The mechanism responsible for topo II-cleavable complex formation by etoposide is now well defined. Etoposide stimulates the ability of topo II to break double-stranded DNA after strand passage of the duplex (Robinson and Osheroff, 1991), and the stabilized complexes then result in a reversible inhibition of religation of the DNA (Sehested and Jensen, 1996). Interestingly, it was recently postulated that it is the interaction between topo II and etoposide (rather than between etoposide and DNA) that directs the formation of cleavable complexes at specific sequences, resulting in inhibition of religation (Burden et al., 1996). In addition, etoposide-induced DNA damage has been shown to trigger apoptosis in many cell lines (e.g., Kamesaki et al., 1993), and the role of p53 in determining cell death is particularly important in cancer cells (Chresta et al., 1996; Skladanowski and Larsen, 1997).

In another study, in which cells treated with VM-26 (another cleavable complex-forming epipodophyllotoxin), topo II-cleavable complexes were immunoprecipitated with isoform-specific antibodies. The results obtained demonstrated that topo II-cleavable complexes were associated with the replication fork (i.e., nascent DNA), whereas topo II complexes were found in bulk (mature) DNA (Qiu et al., 1996). These results highlight the requirement to differentiate between isoforms and their known cell-cycle dependence when designing drug regimens.

One of the most important currently used assays for detection of etoposide-induced DNA damage is alkaline elution, a technique that has yielded much information about double-, single-, and protein-linked strand break levels and their repair (Long et al., 1985; Caldecott et al., 1990). However, this process requires radiolabeling of cells followed by elution of DNA from filters; it does not distinguish between isoforms and is unsuitable for small (e.g., patient) cell samples. Methods to measure topo II catalytic activity in patient samples (e.g., Cattan et al., 1996) are available, but again, these do not identify isoform-specific differences. The novel immunological technique described here can detect isoform-specific antibody interactions, which directly reflect cleavable complex formation in individual cells. In combination with data generated using assays that reveal information about events downstream of cleavable complex formation, this assay should assist in providing a wider understanding of the mechanism of action of topo II-targeting drugs. Work in progress with other topo II-targeting agents that form cleavable complexes indicates that it is possible to detect cleavable complex formation in CCRF-CEM cells after treatment with these drugs.

The advantage of being able to study intercell heterogeneity for both  and  isoforms in cell lines and clinical samples should facilitate better chemotherapeutic regimens. In the light of the data shown in Fig. 6, it will be interesting to measure the disappearance of cleavable complexes in pairs of sensitive and resistant cells. It has been shown that some etoposide-resistant cells have a defect in cleavable complex stability, allowing them to repair more quickly (Ritke et al., 1994). It is possible that the level of complex formation/rate of disappearance (which may differ between isoforms) in resistant cells would provide a useful prognostic marker that could be assessed using this immunofluorescence assay. In patients, "low dose" (<1500 mg/m²) etoposide results in a serum concentration of ~2 µM, whereas "high dose" (>4000 mg/m²) results in a concentration of ~10 µM (Joel, 1996). Because these doses are at the lower end of the range used for our in vivo experiments, it may be possible to use this technique to investigate the resistance status of patients receiving this type of treatment. For example, a patient sample containing a significant subpopulation of resistant cells would give a very low integrated fluorescence signal that might predict a poor response to therapy. Indeed, work is currently in progress to quantify topo II- and -cleavable complex formation in samples from leukemia patients undergoing etoposide therapy.