CX-5461 Preferentially Induces Top2α-Dependent DNA Breaks at Ribosomal DNA Loci

Abstract

While genotoxic chemotherapeutic agents are among the most effective tools to combat cancer, they are often associated with severe adverse effects caused by indiscriminate DNA damage in non-tumor tissue as well as increased risk of secondary carcinogenesis. This study builds on our previous work demonstrating that the RNA Polymerase I (Pol I) transcription inhibitor CX-5461 elicits a non-canonical DNA damage response and our discovery of a critical role for Topoisomerase 2α (Top2α) in the initiation of Pol I-dependent transcription. Here, we identify Top2α as a mediator of CX-5461 response in the murine Eµ-Myc B lymphoma model whereby sensitivity to CX-5461 is dependent on cellular Top2α expression/activity. Most strikingly, and in contrast to canonical Top2α poisons, we found that the Top2α-dependent DNA damage induced by CX-5461 is preferentially localized at the ribosomal DNA (rDNA) promoter region, thereby highlighting CX-5461 as a loci-specific DNA damaging agent. This mechanism underpins the efficacy of CX-5461 against certain types of cancer and can be used to develop effective non-genotoxic anticancer drugs.

Keywords: CX-5461; DNA damage pathway; RNA Polymerase I; double-strand breaks; nucleolar surveillance pathway; p53; ribosome biogenesis; topoisomerase.

Figure 1

Top2α mediates resistance to CX-5461 in mice. (A) Schematic of Top2α mutations. The mouse Top2α gene is shown with the N-terminal ATPase, catalytic TOPRIM, DNA-binding winged-helix domain (WHD), and C-terminal domain (CTD) domains highlighted. Seven different Top2α mutations were identified by whole exome sequencing (WES) of DNA extracted from tumors that had acquired resistance to CX-5461 in vivo. Frameshift, missense and nonsense mutations are denoted by stars, circles, and hexagons, respectively. This figure is adapted from Wendorff et al. [40]. (B) Top2α mRNA levels in T2A^WT and T2A^+/K1266 cells. Basal Top2α mRNA abundance in the T2A^WT and T2A^+/K1266 cells was measured by real-time quantitative PCR (qPCR) and normalized to B2M expression. Error bars represent standard deviation (* p < 0.05, n = 3). (C) Top2α protein expression in T2A^WT and T2A^+/K1266 cells. Representative Western blot of basal Top2α protein. (D) The representative image of decatenation assays demonstrating that T2A^WT cells have greater cellular Top2 activity than T2A^+/K1266 cells. The nuclear extracts containing Top2 enzymes are isolated from each group of cells and are serially diluted before incubation with kDNA. In both instances shown, the Top2 activity as measured by the decatenated fraction of each lane is greater in the lower dilutions of the T2A^WT cells than the T2A^+/K1266 cells. (E) CX-5461 in vitro dose–response assay. T2A^WT (green line) and T2A^+/K1266 (red line) cells were treated with 0.3 nM–1 μM CX-5461 for 24 h in triplicate, and the live cell number was counted by volumetric FACS of propidium iodide (PI)-negative cells. Results were normalized to the number of live cells after drug vehicle treatment (100%), and the line of best fit was plotted. Concentrations of the drug inhibiting 50% of growth (GI₅₀) were calculated using GraphPad Prism 10. (F,G) The Kaplan–Meier curve of mice transplanted with T2A^WT (F) and T2A^+/K1266 (G) cells ± CX-5461. Mice were treated every three days with 35 mg/kg of CX-5461. Median survival: T2A^WT = 15 days, T2A^+/K1266 = 1 day (*** p < 0.001, n.s. = not significant, n = 8/group).

Figure 2

Top2α determines the level of activation of DDR pathways in CX-5461-treated cells but not the ability of CX-5461 to inhibit Pol I transcription. (A) T2A^WT and T2A^+/K1266 cells were incubated with radiolabeled orthophosphate for 20 min and then treated with varying concentrations of CX-5461 as indicated. RNA was isolated and analyzed on agarose-formaldehyde gel. Newly synthesized 47 pre-rRNA was detected by autoradiography (a representative experiment is shown; upper panel) and total 28S rRNA was detected by ethidium bromide staining (a representative experiment is shown; bottom panel). (B) Representative Western blots showing the downstream signaling response to CX-5461 ± KU-55933 treatment of T2A^WT and T2A^+/K1266 cells. Cells were treated with 30 nM CX-5461 for 30–180 min ± 30 min pre-treatment with ATM inhibitor KU-55933. Western blots were probed for phosphorylated ATM (S1981), phosphorylated p53 (S15), total p53 and loading controls tubulin and β-actin. (C,D) Quantitative real-time PCR of (C) p21 and (D) Puma from the experiment outlined in (B). Error bars represent standard deviation (* p < 0.05, ** p < 0.01, *** p < 0.001, n = 5). Dotted line represents average relative expression value at time point 0. (E) T2A^WT and T2A^+/K1266 cells were treated with varying concentrations of CX-5461 (as in Figure 1E) for 24 h after pre-treatment with ±1 μM KU-55933 for 30 min in triplicate. Cell viability was determined by measuring absorption at 570 nm after incubation with Alamar Blue. Results were normalized to vehicle-treated cells (set 100%), and GI₅₀ values were determined and plotted as a bar graph. Error bars represent standard deviation (n = 3).

Figure 3

CX-5461 does not induce significant formation of Top2α (Top2α-cc) cleavage complexes, but causes low-level, nucleolar-localized DNA damage. (A) Schematic of the TARDIS assay. Cells were suspended in a thin layer of agarose on a microscope slide. After lysis with SDS and incubation with a high salt buffer, only DNA and macromolecules covalently bound to the DNA remain on the slide. Covalently bound proteins, such as active Top2α can be imaged by immunofluorescent labelling. (B) Representative images of Eμ-Myc T2A^WT cells pre-treated with 1 μM MG-132 for 15 min to prevent proteolytic degradation of DNA-bound Top2α, then treated with drug vehicle (top panel), 30-nM CX-5461 (middle panel), or 50-nM doxorubicin (bottom panel) for 2 h. The DNA is stained blue by DAPI, and Top2α bound covalently to DNA is shown in green. Scale bar, 5 μm (C) Quantitation of the experiment is outlined in (B). While CX-5461 exhibits a similar degree of Top2 inhibition via the decatenation assay (Figure S3), CX-5461 causes significantly fewer Top2α-DNA complexes in cells than doxorubicin (n = 45–47 cells, *** p < 0.001). (D) Representative images of Eμ-Myc T2A^WT cells treated with either drug vehicle, 30 nM CX-5461, or 60nM Etoposide for 1 h immunostained for the nuclei (DAPI; blue), the nucleoli (Fibrillarin; green) and γH2AX (red). ). Scale bar, 10.4 μm. (E) Quantitation of the distance to the closest fibrillarin spot from each γH2AX foci. Whiskers extend to the 10th and 90th percentiles, and the mean is represented by the cross. CX-5461-treated cells exhibit a shorter median γH2AX-Fib distance compared to vehicle- or etoposide-treated cells (** p < 0.01). (F) Quantitation of the proportion of sub-0.3 μm γH2AX-Fib distances for both CX-5461 and etoposide-treated samples from three independent experiments. Error bars represent the standard deviation (* p < 0.05).

Figure 4

CX-5461 treatment induces targeted, Top2α-mediated DNA damage at the rDNA loci. (A) HTETOP cells were grown for 24 h either untreated (−Dox) or in the presence of 1 mg/mL Doxycycline (+Dox) and then treated with ±2 μM CX-5461 for 1 h before the DNA double-strand breaks (DSB) were labelled with BrdUTP using terminal transferase. DNA was extracted and sonicated to 400–600 bps fragments, and the labelled DNA was enriched by anti-BrdUTP antibodies. DNA enrichment was quantitated by qPCR. Signal was normalized to control (IgG) and plotted as fold changes. Error bars represent the standard deviation (n = 3). (B) HTETOP cells were grown for 24 h either untreated (−Dox) or in the presence of 1 mg/mL Doxycycline (+Dox) and then treated with ±5 µM etoposide for 2 h. Samples were processed and analyzed as in (A). Error bars represent the standard deviation (n = 3). (C) HTETOP cells were grown for 24 h either untreated (-Dox) or in the presence of 1 mg/mL Doxycycline (+Dox) and then treated with ±3 nM Actinomyci D (ActD) for 4 h. Samples were processed and analyzed as in (A). Error bars represent the standard deviation (n = 3). (D) HTETOP cells were grown for 24 h untreated (no Dox) and then treated with 2 μM CX-5461 for different period of time as indicated. Samples were processed and analyzed as in (A). Error bars represent the standard deviation (n = 3).