Enhanced etoposide sensitivity following adenovirus-mediated human topoisomerase II α gene transfer is independent of topoisomerase II β

The roles that the α and β isoforms of topoisomerase II (topo II) play in anticancer drug action were determined using MDA-VP etoposide-resistant human breast cancer cells and a newly constructed adenoviral vector containing the topo IIα gene (Ad-topo IIα). MDA-VP cells were more resistant to etoposide than to amsacrine and had more resistance to etoposide than did MDA-parental cells. MDA-VP cells also expressed lower topo IIα RNA and protein levels than parental cells but had comparable topo IIβ levels. After infection with Ad-topo IIα, topo IIα, RNA and protein levels increased significantly, as did the cells' sensitivity to etoposide. In contrast, topo IIβ levels remained constant with little alteration in the cells' sensitivity to amsacrine. Band-depletion immunoblotting assays indicated that topo IIα was depleted in etoposide-treated, Ad-topo IIα-transduced MDA-VP cells but not in amsacrine-treated cells. Topo IIβ was depleted in amsacrine-treated, Ad-topo IIα-MDA-VP cells, with little change in the topo IIα levels. These results suggest that topo IIα gene transfer does not alter topo IIβ expression and that enhanced sensitivity to etoposide is therefore secondary to change in topo IIα levels. These studies support the theory that etoposide preferentially targets topo IIα, while amsacrine targets topo IIβ. © 2001 Cancer Research Campaign http://www.bjcancer.com

MDA-MB-231 parental cells were obtained from American Type Culture Collection (Manassas, VA, USA). MDA-VP etoposideresistant human breast cancer cells were initially derived and cloned from MDA-parental cells as described previously (Matsumoto et al, 1997). All cells were screened and found to be free of Mycoplasma (Gen-Probe Co., San Diego, CA, USA).

Infection of cells with Ad-topo IIα virus
The Ad-topo IIα virus was constructed and purified as described previously (Zhou et al, 1999). Cells were grown in logarithmic phase and were infected with Ad-topo IIα at a multiplicity of infection of 100 pfu cell. Cells were harvested by standard methods after 48 h.

Band-depletion immunoblotting assay
The band-depletion immunoblotting assay was performed as described previously (Zwelling et al, 1989). Cells were infected with Ad-topo IIα or Ad-β-gal (control) for 48 h and then treated with 200 µM etoposide or 100 µM amsacrine at 37˚C for 1 h as indicated. Cell lysates were prepared in 2X Laemmli buffer by sonication for 30 s and boiled for 5 min. Proteins were resolved on a 7.5% SDS/polyacrylamide gel and immunoblotted using human topo IIα, topo IIβ, and β-actin antibodies.

RESULTS
MDA-VP and parental cells were treated with various concentrations of etoposide or amsacrine. Table 1 shows that MDA-VP cells were 15-fold more resistant to etoposide than were MDA-parental cells. In contrast, MDA-VP cells were only 2.2-fold more resistant to amsacrine. To determine whether the differences in resistance were related to expression of the topo II isoforms in these cells, topo IIα and topo IIβ RNA and protein levels were measured.
Topo IIα mRNA levels were lower in MDA-VP cells than in MDA-parental cells ( Figure 1A). Densitometric analysis showed only 20% topo IIα gene expression in the etoposide-resistant MDA-VP cells compared to the MDA-parent cells ( Figure 1B,  To further explore the relationship between drug resistance and isoform expression, MDA-VP cells were infected with the Adtopo IIα virus, then the topo II RNA and protein levels, as well as drug sensitivity, were quantified. Topo IIα mRNA levels were elevated after infection; however, topo IIβ levels were not significantly altered (Figure 1). Topo IIα protein levels also increased in MDA-VP cells after Ad-topo IIα infection, but topo IIβ protein levels remained constant (Figure 2). The sensitivity of MDA-VP cells to etoposide increased 4.5-fold after infection with Ad-topo IIα. The IC 50 of MDA-VP cells infected with Ad-topo IIα went from 45.6 µM to 10.1 µM. By contrast, the sensitivity to amsacrine only increased 1.3-fold following infection with Adtopo IIα (Table 2). Therefore, the increased sensitivity of cells to etoposide following topo IIα gene transfer correlated with increased topo IIα levels but not with topo IIβ levels.
A band-depletion immunoblotting assay was performed with topo IIα and topo IIβ antibodies to analyze the interaction of the 2 isoforms with etoposide and amsacrine. Topo IIα band depletion was seen following treatment with etoposide, while little change was seen following amsacrine treatment (Figure 3, lanes  2,3, upper panel). In contrast, topo IIβ was more depleted in cells treated with amsacrine than in cells treated with etoposide ( Figure  3, lanes 2,3, middle panel). Densitometric analysis indicated that etoposide induced 70% depletion of topo IIα protein and only 10% depletion of topo IIβ protein (Figure 3, lane 2). Conversely, amsacrine induced only a 10% reduction of topo IIα protein, but a 60% reduction of topo IIβ protein (Figure 3, lane 3). After infection of MDA-VP cells with Ad-topo IIα, topo IIα protein levels were once again significantly increased (Figure 3, lane 4, upper panel) with relatively no change in topo IIβ protein levels ( Figure  3, lane 4, middle panel). Neither topo IIα protein levels nor topo IIβ protein levels were significantly altered following infection with Ad-β-gal (Figure 3, lane 7). The band-depletion pattern in MDA-VP cells following infection with Ad-topo IIα (Figure 3, lanes 5, 6) and Ad-β-gal ( Figure 3, lanes 8, 9) was the same as that seen in MDA-VP control cells (Figure 3, lanes 2,3). Etoposide treatment induced a significant reduction in topo IIα, with little change in topo IIβ. By contrast, treatment with amsacrine did not affect the increased topo IIα protein levels in the MDA-VP-Adtopo IIα cells.

DISCUSSION
Topo II, a nuclear enzyme involved in a number of important cellular processes, is the target for several anticancer drugs. The specific roles of topo IIα and topo IIβ isoforms in the action of these topo II-targetting drugs are still poorly understood. Our data provide further evidence that topo IIα is the main target for etopo-  side, while topo IIβ is the preferred target for amsacrine in MDA-VP cells. The etoposide sensitivity and resistance are more related to topo IIα gene expression than to topo IIβ expression. MDA-VP cells expressed lower levels of topo IIα RNA and protein than MDA parental cells. In contrast, topo IIβ RNA and protein levels were relatively the same in both cell types. MDA-VP cells are more resistant to etoposide than to amsacrine and this correlates to the topo IIα and topo IIβ protein levels. Correlation between mRNA topo II levels and cell kill are not always universal. The level of drug-stabilized cleavable complex formation is the most important factor (Koo et al, 1999). Our previous studies show that etoposide-induced topo IIα-DNA cleavable complex formation is also significantly lower in MDA-VP cells than in parental cells, supporting the hypothesis that low levels of topo IIα account for the etoposide resistance of these cells. Drug uptake and participation of P-glycoprotein or the multiple drug-resistant associated protein do not play a role in resistance of MDA-VP cells .
Transfer of the human topo IIα gene into MDA-VP cells using an adenoviral vector increased topo IIα protein levels without an appreciable change in topo IIβ protein levels. The topo IIα protein produced following transduction was sensitive to etoposide but not to amsacrine. Etoposide-induced cytotoxicity was enhanced 4.5fold in cells transduced with topo IIα, whereas amsacrine-induced cytotoxicity did not change significantly. These results indicate that topo IIα gene transfer does not alter topo IIβ expression and that the enhanced sensitivity to etoposide is secondary to the change in topo IIα.
The involvement of topo IIβ in amsacrine sensitivity is also supported by others. Herzog et al (1998) have shown that topo IIβ mRNA levels in HL60/AMSA amsacrine-resistant human leukaemia cells are only 10% of those in HL-60 parental cells and that topo IIβ protein is not detectable in HL60/AMSA cells. However, these cells are sensitive to etoposide. Withoff et al (1996) have additionally demonstrated that amsacrine resistance in GLCA/AM3y cells, a subline of the human small cell lung carcinoma cell line, is linked to a major decrease in topo IIβ protein. Dereuddre et al were able to increase the sensitivity of a Chinese hamster lung cell line to amsacrine by transfection with the topo IIβ gene (Dereuddre et al, 1997).
Topo IIβ have different tissue distribution. High levels of topo IIα expression have been seen in aggressive proliferating tumours, whereas topo IIβ appears to be expressed ubiquitously in quiescent cells (Turley et al, 1997). Topo IIα is essential for survival of eucaryotic cells (Wang, 1996), while topo IIβ does not appear to be essential for either proliferation or survival (Yang et al, 2000;Herzog et al, 1998). Such findings may help explain the greater clinical utility of etoposide versus amsacrine. Each topo II isoform appears to carry out a different cellular function and plays a different role in drug resistance. It is important to understand how tumour cell sensitivity may be influenced by differential expression of these two isoforms.
In summary, our data indicate that topo IIα gene transfer does not affect topo IIβ expression, and the ability to circumvent etoposide resistance using topo IIα gene transfer is secondary to enhanced production of the drug-sensitive protein. These data substantiate the hypothesis that etoposide preferentially targets topo IIα, while amsacrine targets topo IIβ. In addition we have shown that we can successfully manipulate topo IIα gene expression in cells without the problem of feedback inhibition previously experienced by us and other laboratories . We attributed this to our use of the strong cytomegalovirus promoter in our adenoviral vector construct. This vector can be manipulated by making mutations in specific parts of the gene and thus provide a valuable tool with which to investigate the biology of human topo IIα expression and function.  6) or Ad-β-gal (lanes 7-9). Topo II protein isoforms were extracted and quantified using a band-depletion immunoblotting assay with human topo IIα, topo IIβ, and β-actin antibodies. The relative fold was calculated using densitometric analysis from 3 independent experiments. MDA-VP cells were designated as 1.0 and calculations were adjusted according to the β-actin protein loading control University of Medicine and Dentistry of New Jersey, Piscataway, NJ, USA for the generous gift of human topo IIα gene probe ZII69. We also wish to thank Dr Caroline Austin, University of Newcastle-upon-Tyne, Newcastle, UK for the gift of topo IIβ gene probe F12. This work was supported in part by NIH Grants CA42992 (to ESK), CA40090 (to LAZ), and Cancer Center Support Core Grant CA16672 (to MD Anderson Cancer Center).