Toward discovering new anti-cancer agents targeting topoisomerase IIα: a facile screening strategy adaptable to high throughput platform

Abstract

Topoisomerases are a family of vital enzymes capable of resolving topological problems in DNA during various genetic processes. Topoisomerase poisons, blocking reunion of cleaved DNA strands and stabilizing enzyme-mediated DNA cleavage complex, are clinically important antineoplastic and anti-microbial agents. However, the rapid rise of drug resistance that impedes the therapeutic efficacy of these life-saving drugs makes the discovering of new lead compounds ever more urgent. We report here a facile high throughput screening system for agents targeting human topoisomerase IIα (Top2α). The assay is based on the measurement of fluorescence anisotropy of a 29 bp fluorophore-labeled oligonucleotide duplex. Since drug-stabilized Top2α-bound DNA has a higher anisotropy compared with free DNA, this assay can work if one can use a dissociating agent to specifically disrupt the enzyme/DNA binary complexes but not the drug-stabilized ternary complexes. Here we demonstrate that NaClO4, a chaotropic agent, serves a critical role in our screening method to differentiate the drug-stabilized enzyme/DNA complexes from those that are not. With this strategy we screened a chemical library of 100,000 compounds and obtained 54 positive hits. We characterized three of them on this list and demonstrated their effects on the Top2α-mediated reactions. Our results suggest that this new screening strategy can be useful in discovering additional candidates of anti-cancer agents.

Figure 1. A schematic diagram of fluorescence-based screening for Top2α targeting compounds.

(A) The fluorescent DNA substrate was synthesized with a known Top2 cleavage site (indicated by arrows). Alexa Fluor 488 marked by asterisk was labeled at the 3′end. (B) Schematic representation of the anisotropy-based Top2α drug screening. The topoisomerase-DNA complex, covalent or non-covalent, has a slower rotational diffusion and results in a higher anisotropy. After treated with a protein-dissociating agent, the non-covalent bound Top2α was removed, and the free DNA has a lower anisotropy. If a Top2α-targeting agent stabilized covalent enzyme-DNA complexes, the anisotropy will remain high.

Figure 2. Dose-dependent anisotropy from VM26 after treatment with NaClO₄.

(A) Reactions of Alexa Fluor 488-labeled DNA with human Top2α were carried out with VM26 concentrations of 60, 120, 180, 240 and 300 µM. A dose-dependent increase in anisotropy with increasing VM26 concentrations was observed if the reactions were terminated with NaClO₄ (solid square) but not without it (empty triangle). (B) The dose-dependent increase of anisotropy with increasing VM26 was observed when the reactions contained human Top2α (solid square) but not without it (empty square).

Figure 3. NaClO₄ abrogates Top2α binding to DNA, but retains the drug-stabilized Top2α-DNA ternary complex.

(A) The VM26-dependent increase in anisotropy was observed when the reaction was terminated with NaClO₄ (Reaction IV vs Reaction III), and not when NaClO₄ was added in the presence of Top2α before the addition of DNA, regardless of the addition of VM26 (Reactions VI and VII vs Reaction IV). Anisotropy observed in the controls was DNA only (Reaction I), DNA/Top2α without VM26 (Reaction II), and incubation without Mg⁺⁺ (Reaction V). (B) Optimization of NaClO₄ concentration for anisotropy assay. The reaction was performed in 50 µl Top2 reaction buffer with 300 µM VM26, 50 nM Alexa Fluor 488-labeled DNA and 250 nM human Top2α. The reaction mixture was incubated at 37°C for 30 minutes. Before measuring the anisotropy, NaClO₄ was added into each reaction mixture to give a final concentration up to 750 mM. The maximal extent of dissociation was reached at a concentration of 100 mM or above. (C) Anisotropy observed with etoposide (VP16), mAMSA, mitoxantrone (MXT), and with teniposide (VM26) as positive controls, after treatment with 200 mM NaClO₄. The concentrations of these known Top2α drugs used here were 300 µM for VP16 and VM26, and 15 µM for mAMSA and MXT.

Figure 4. NaClO₄ induces the formation of cleavage complexes with a preference of nicked DNA.

(A) Reactions of human Top2α and pUC19 plasmid DNA were terminated with either NaClO₄ or SDS. Addition of NaClO₄ to 100 mM leads to the formation of more nicked DNA products while the addition of SDS results in a preference of linear ones. Reactions were carried out with 8.5 nM pUC19, 3.75 nM Top2α, and 100 µM VM26, terminated with the addition of either NaClO₄ or SDS, the reaction products were treated with proteinase K, and analyzed by electrophoresis in 1.2% agarose gel containing 0.15 µg/ml ethidium bromide. (B) DNA cleavage assays were performed under similar conditions except with increasing concentrations of VM26 followed by addition of NaClO₄. An increase in the nicked DNA cleavage products was observed with higher amounts of VM26. (C) Plasmid DNA cleavage assays were carried out using different enzyme to DNA ratios as indicated. Linear DNA products (full-length or sub-full length) from SDS treatment, and both nicked and linear DNA from NaClO₄ treatment increase with higher enzyme to DNA ratios. (D) Reactions in the presence of various concentrations of VM26 contained linear pUC19 as substrate with an enzyme/DNA ratio of 20, followed by adding NaClO₄ or SDS, or both sequentially, to stop the reactions. The size markers, and linear pUC19 were loaded in the leftmost two lanes. Positions of DNA were marked as NC (nicked circular), L (linear), NSC (negative supercoiled), and RC (relaxed).

Figure 5. Glycerol gradient sedimentation analysis of topoisomerase/DNA complexes.

The glycerol gradient solution is layered in a polyallomer tube starting from 60% to 30% glycerol. The reaction for Top2α/DNA complex formation was performed in 80 µl of a solution containing 100 µM VM26, 34 nM pUC19 and 3.75 nM human Top2α. The reaction mixture was loaded on top of a 30–60% glycerol gradient, and tube was topped off with mineral oil. After ultracentrifugation, the fractions were collected from the bottom of the tubes. The distribution of Top2α in each fraction was analyzed by slot blotting and immunodetection.

Figure 6. Assay development of high throughput screening.

For development of HTS assay, we carried out reactions in 1536-well microtiter plate with various enzyme/DNA ratios in 4 µl of reaction buffer containing 10 mM Tris-HCl pH 7.9, 5 mM MgCl₂, 50 mM NaCl, 0.1 mM EDTA, 0.5 mM ATP with or without 300 µM VM26. In the reaction mixture, the enzyme/DNA ratio started with 0.75 (150 nM enzyme, and 200 nM DNA) and increased by two-fold up to a ratio of 6. The reaction mixture was incubated at 37°C for 30 minutes. Before measuring the anisotropy, NaClO₄ was added into each reaction mixture to give a final concentration of 100 mM. Z factor was calculated according to Zhang et al . Anisotropy was measured in a unit of mA (milliAnisotropy).

Figure 7. HTS process diagram and outcomes.

(A) Diagrammatical representation of HTS procedure with 1536-well microtiter plates. The detailed screening protocol is described in Materials and Methods. (B) Distribution histogram of activities assayed. Activity is determined by taking the anisotropy ratio of the test compound over the mock controls. The ordinate is for the number of compounds screened. To highlight the bimodal distribution, red portion is for 10× compound counts of green portion.

Figure 8. Confirmation of the positive hits.

(A) Chemical structure of VM26 and three positive hits, betulinic acid, DDI and CET. (B) Anisotropy of enzyme/DNA/drug ternary complexes. The reaction was performed in 50 µl of reaction buffer with 50 nM Alexa Fluor 488-labeled DNA and 250 nM human Top2α. Negative control (filled) contained neither VM26 nor test compounds. Positive control (open) contained 300 µM VM26. Others (gray) contained 15 µM of test compounds. The reaction mixture was incubated at 37°C for 30 minutes. Before measuring the anisotropy, NaClO₄ was added into the reaction mixture to give a final concentration of 200 mM. (C) Effects of positive hits on inhibition of DNA relaxation. The reaction contained 8.5 nM pUC19 and 0.125 nM human Top2α in 20 µl reaction buffer, and performed at 37°C for 5, 10 or 15 minutes. Solvent control contained 5% DMSO, and the other reactions contained 60 µM hit compound (DMSO was used as the solvent and its final concentration in each reaction was 5 %). (D) Effects of positive hits on DNA cleavage assay. The 20 µl reaction contained 8.5 nM pUC19 and 3.75 nM human Top2α with Top2α targeting compounds at the concentrations marked above each lane. Reaction products were analyzed by electrophoresis in a 1.2% agarose gel containing 0.15 µg/ml ethidium bromide. Positions of DNA were marked as NC (nicked circular), L (linear), NSC (negative supercoiled), and RC (relaxed).

Figure 9. Cytotoxicity of the positive hits.

HCT116 cells were treated with DDI, CET, and teniposide (VM26) at various concentrations, and their cytotoxicity was determined by MTS assays. The cytotoxicity toward teniposide is highlighted in the insert.