Reduced expression of DNA topoisomerase I in SF295 human glioblastoma cells selected for resistance to homocamptothecin and diflomotecan

Abstract

Homocamptothecins (hCPTs) are a novel class of topoisomerase I (Top1) inhibitors with enhanced chemical stability compared with the currently used camptothecin (CPT) analogs irinotecan and topotecan. The hCPT derivative diflomotecan (BN80915) is currently in clinical trials. We established two resistant human glioblastoma cell lines, SF295/hCPT50 and SF295/BN50, by stepwise exposure of the parental SF295 line to increasing concentrations of hCPT and BN80915, respectively. The two resistant cell lines were 15- to 22-fold resistant to hCPT and BN80915 as well as 7- to 27-fold cross-resistant to other Top1 inhibitors, including CPT, topotecan, and the indenoisoquinolines MJ-III-65 (NSC 706744) and NSC 724998, but sensitive to the topoisomerase II inhibitors mitoxantrone and etoposide. Neither of the resistant cell lines displayed any detectable expression of the three major drug transporters P-glycoprotien, multidrug resistance-associated protein 1, or ATP-binding cassette, subfamily G (WHITE), member 2, as assessed by immunoblot or flow cytometry. Reduced expression of Top1 protein occurred at the transcriptional level in both of the resistant cell lines, consistent with the reduction of Top1 enzyme level as the major contribution to the resistance phenotype in SF295/hCPT50 and SF295/BN50 cells. Treatment of the resistant cell lines with the histone deacetylase inhibitor depsipeptide or the DNA methyltransferase inhibitor 5-aza-2'-deoxycytidine alone or concomitantly did not result in re-expression of Top1. Our studies suggest that selection for resistance to hCPT or BN80915 is primarily related to reduced Top1 expression at the transcriptional level, resulting in reduced enzyme levels.

Fig. 1

Chemical structures of camptothecin, homocamptothecin, and diflomotecan (BN80915).

Fig. 2

Analysis of drug transporter protein expression in human glioblastoma SF295 and drug-resistant SF295/BN50 and SF295/hCPT50 cells. A, for qualitative analysis of the three drug transporter proteins expression in these cell lines, 30 μg of microsomal proteins were separated by SDS-PAGE and probed with anti-Pgp, -MRP1, and -ABCG2 antibodies as described under Materials and Methods. B, Ponceau staining of membrane shown in A. C, SF295 parental and resistant sublines from A were incubated with 0.5 μg/ml rhodamine 123 in the presence (dashed line) or absence (solid line) of 3 μg/ml valspodar (column 1); 200 nM calcein AM in the presence (dashed line) or absence (solid line) of 50 μM MK571 (second column); or 10 μM pheophorbide A in the presence (dashed line) or absence (solid line) of 10 μM FTC (last column) to detect Pgp, MRP1, or ABCG2, respectively. ABCB1-, ABCC1-, or ABCG2-transfected HEK293 cells served as positive controls for Pgp, MRP1, and ABCG2, respectively (+ Control, top row of histograms). Representative results from one of two independent experiments are shown.

Fig. 3

DNA-protein crosslinks (DPCs) induced by CPT in human glioblastoma SF295 and BN80915-resistant SF295/BN50 cells. A, cells were prelabeled with [³H]thymidine and were then treated with 0.1, 0.3, and 1.0 μMCPT for 1 h at 37°C as described under Materials and Methods. DPCs were assayed by alkaline elution under nondeproteinizing conditions after ionizing radiation of cells with 3000 rads for reference (squares); or treatment with 0.1 μM CPT (circles); 0.3 μM CPT (triangles) or 1.0 μM CPT (diamonds) in SF295 (closed symbols) and SF295/BN50 (opened symbols) cells. B, DPCs were measured by alkaline elution assay and are expressed as rad equivalents at 12 h elution time in SF295 (◯) and SF295/BN50 (◻) cells. The average DPCs from two independent experiments are presented. Bars, S.E.M.

Fig. 4

Top1 (A), Top2α (B), and Top2β (C) protein levels by Western blotting analysis, Top1 mRNA levels by real time quantitative-PCR analysis (D), and effect of decitabine (1 μM) and depsipeptide (2 ng/ml), alone or in combination, on mRNA expression of Top1 by semiquantitative RT-PCR analysis (E) in SF295, SF295/BN50, and SF295/hCPT50 human glioblastoma cells. For Western blotting assay, whole-cell lysates (50 μg) isolated from each cell line were probed with monoclonal mouse Top1 and polyclonal Top2 antibodies as described under Materials and Methods. β-actin was probed to show equal loading. For real-time quantitative-PCR analysis, total RNA was extracted, and real-time quantitative-PCR was performed as described under Materials and Methods. mRNA levels were normalized with 18S RNA and representative of at least two independent experiments. For semiquantitative RT-PCR analysis of Top1 in SF295 parental cells and the respective resistant sublines SF295/BN50 and SF295/hCPT50, cells were untreated or treated with either decitabine (1 μM) for 4 days, depsipeptide (2 ng/ml) for 1 day, or a combination of the two drugs. Representative results from three independent experiments are shown. The numbers indicates the -fold difference of Top1 relative to the untreated SF295 parental cells after normalization to GAPDH.

Fig. 5

BrdU incorporation and cell cycle analysis in SF295, SF295/BN50, and SF295/hCPT50 cells. A, cells were collected at indicated times after BrdU pulse, and stained with anti-BrdU antibody coupled to FITC and propidium iodide/RNase A. B, quantification of BrdU-positive cells that are still in the S-phase at indicated times after the BrdU pulse. The first time point (3 h after the BrdU pulse) is set up at 100%. The average percentages are presented from four independent experiments. Bars, S.E.M.