Poly(ADP-ribose) polymerase inhibitor induces accelerated senescence in irradiated breast cancer cells and tumors

Abstract

Persistent DNA double-strand breaks (DSB) may determine the antitumor effects of ionizing radiation (IR) by inducing apoptosis, necrosis, mitotic catastrophe, or permanent growth arrest. IR induces rapid modification of megabase chromatin domains surrounding DSBs via poly-ADP-ribosylation, phosphorylation, acetylation, and protein assembly. The dynamics of these IR-induced foci (IRIF) have been implicated in DNA damage signaling and DNA repair. As an IRIF reporter, we tracked the relocalization of green fluorescent protein fused to a chromatin binding domain of the checkpoint adapter protein 53BP1 after IR of breast cancer cells and tumors. To block DSB repair in breast cancer cells and tumors, we targeted poly(ADP-ribose) polymerase (PARP) with ABT-888 (veliparib), one of several PARP inhibitors currently in clinical trials. PARP inhibition markedly enhanced IRIF persistence and increased breast cancer cell senescence both in vitro and in vivo, arguing for targeting IRIF resolution as a novel therapeutic strategy.

Figure 1

GFP-IBD reporter reveals IRIF formation and kinetics. A, shRNA knockdown of the upstream DNA damage response protein ATM, MDC1 or RNF8 abrogates GFP-IBD reporter relocalization to IRIF. shRNA knockdown of endogenous 53BP1 increases foci numbers at 3 and 24 h. Scale bar, 10 μm. B, Formation and resolution of GFP-IBD foci in response to IR depend on time and dose. Scale bar, 10 μm. C, Mean number of GFP-IBD foci per cell ± SD (n > 50) at increasing IR doses evaluated at 3 h (solid circles) and 24 h (solid squares). D, Clonogenic survival of MCF7^Tet-On GFP-IBD cells treated with increasing doses of IR (means ± SD, n = 3). Clonogenicity is modeled as distinct regimes of lower lethality from 0 to 6 Gy (% survival = 330 × e^{−0.89 x (dose in Gy)}, R² = 0.958) and higher lethality from 6 to 12 Gy (% survival = 20.0 × e^{−1.99 x (dose in Gy)}, R² = 0.999). Clonogenic efficiency of untreated MCF7^Tet-On GFP-IBD cells represents 100% control.

Figure 2

PARP1 inhibitor ABT-888 (veliparib) alters IRIF dynamics and suppresses cell proliferation. A, ABT-888 increased the number of residual IRIF 24 h after IR. Cells were pretreated with DMSO (control) or 10 μM ABT-888 (inset) for 30 min before IR. Live cell images shown at 3 h and 24 h. Scale bar, 5 μm. B, Immunofluorescence reveals co-localization of GFP-IBD with γH2AX, MDC1, and endogenous, full-length 53BP1 at IRIF in cells treated with 6 Gy ± ABT-888. Nuclei indicated by DAPI staining (blue). Scale bar, 10 μm. C, Time-lapse live cell imaging of GFP-IBD localization in MCF7^Tet-On GFP-IBD cells treated with 6 Gy ± ABT-888. Scale bar, 10 μm. D, ABT-888 suppresses cell growth of irradiated MCF7^Tet-On GFP-IBD cells. Cells were treated as shown, fixed at 10 d and stained with crystal violet.

Figure 3

ABT-888 induces accelerated senescence in irradiated MCF7^Tet-On GFP-IBD cells. A, MCF7^Tet-On GFP-IBD cells treated with IR + ABT-888 display persistent IRIF and develop senescent morphology. Scale bar, 20 μm. B, IR + ABT-888 increases SA-βGal activity. Cells were treated with 6 Gy ± ABT-888, fixed at 7 d and stained. Scale bar, 20 μm. C, Relative p21^Cip1/WAF1 transcript levels after IR ± ABT-888 determined by qPCR (mean ± SD, IR, white bars; IR + ABT-888, grey bars). D, Immunofluorescence staining for p21^Cip1/WAF1 at 4 d after 3 Gy ± ABT-888. Scale bar, 20 μm.

Figure 4

IR + ABT-888 induces persistent IRIF and senescence in vivo and suppresses MCF7^Tet-On GFP-IBD tumor regrowth. A, Dose-response of IRIF formation in xenograft tumor cells 24 h after IR. Scale bar, 10 μm. B, IR + ABT-888 increases residual IRIF compared to IR alone. Intravital imaging of GFP-IBD foci in tumors at 3 and 24 h after 3 Gy ± ABT-888. Scale bar, 10 μm. C, SA-βGal activity in tumors treated with IR ± ABT-888. Frozen sections of excised tumors 7 d after IR were fixed and stained. Scale bar, 20 μm. D, ABT-888 + IR suppresses tumor regrowth. Tumor growth was significantly delayed after 6 Gy + ABT-888 (solid squares) compared to 6 Gy (solid circles). Data graphed as mean fractional volume (V/V_o) ± SEM (n = 4/group).