HDAC inhibitor PCI-24781 decreases RAD51 expression and inhibits homologous recombination

Abstract

Histone deacetylase (HDAC) inhibitors such as the phenyl hydroxamic acid PCI-24781 have emerged recently as a class of therapeutic agents for the treatment of cancer. Recent data showing synergy of HDAC inhibitors with ionizing radiation and other DNA-damaging agents have suggested that HDAC inhibitors may act, in part, by inhibiting DNA repair. Here we present evidence that HDAC enzymes are important for homologous recombinational repair of DNA double-strand breaks. Combination studies of PCI-24781 with the poly(ADP-ribose) polymerase inhibitor PJ34, an agent thought to produce lesions repaired by homologous recombination (HR), resulted in a synergistic effect on apoptosis. Immunofluorescence analysis demonstrated that HDAC inhibition caused a complete inhibition of subnuclear repair foci in response to ionizing radiation. Mechanistic investigations revealed that inhibition of HDAC enzymes by PCI-24781 led to a significant reduction in the transcription of genes specifically associated with HR, including RAD51. RAD51 protein levels were significantly decreased after 24 h of drug exposure both in vitro and in vivo. Consistent with inhibition of HR, treatment with PCI-24781 resulted in a decreased ability to perform homology directed repair of I-SceI-induced chromosome breaks in transfected CHO cells. In addition, an enhancement of cell killing was observed in Ku mutant cells lacking functional nonhomologous end joining compared with WT cells. Together these results demonstrate that HDAC enzymes are critically important to enable functional HR by controlling the expression of HR-related genes and promoting the proper assembly of HR-directed subnuclear foci.

Fig. 1.

Synergy of HDAC and PARP inhibition. HCT116 cells were treated with either PCI-24781 or PJ34 alone or in combination at the indicated doses. After treatment for 96 h, cell toxicity was assayed by Annexin V–FITC staining and quantitated by flow cytometry. The percentage of apoptosis is represented on the ordinate. The combined effect of both drugs was quantitatively evaluated by the median effects method as described in Materials and Methods. A CI of <1.0 was obtained, indicating strong synergism.

Fig. 2.

Inhibition of subnuclear repair foci by PCI-24781. Shown are immunofluorescence images of HCT116 cells stained with anti-RAD51 antibody after 24 h of pretreatment with PCI-24781 and at either 1 h (A) or 16 h (B) after irradiation (IR). DAPI shows nuclear staining.

Fig. 3.

PCI-24781 treatment decreases transcript levels of HR genes. The effect of PCI-24781 (0.2 μM) treatment for various times on the transcript levels of selected DNA-repair-associated genes [BRCA1 (A), BRCA2 (B), RAD51 (C), and GADD45γ (D)] in HCT116 cells was quantified by TaqMan analysis and is shown relative to the DMSO control.

Fig. 4.

PCI-24781 decreases RAD51 protein levels. (A) Effect of PCI-24781 on RAD51 protein expression, acetylated tubulin, and acetylated histones for 6 or 24 h was evaluated in HCT116 cells by Western blotting. (B) RAD51 down-regulation is not dependent on caspase activity. RAD51 and full-length and cleaved PARP protein levels were assessed by Western blotting after a 1-h pretreatment of HCT116 cells with 10 μM Q-VD-OPh, a pan-caspase inhibitor, and 24-h treatment with the indicated doses of PCI-24781. (C) An oral 200 mg/kg dose of PCI-24781 reduces RAD51 in HCT116 tumor xenografts in vivo. RAD51, acetylated tubulin, and actin levels after extraction of tumor from HCT116 mouse xenografts with different dosing regimens are shown. 1X animals received a single oral dose 4 h before the end of the study; 2X animals received one oral dose 28 h before the end of the study and received a second dose 6 h later; 3X animals were dosed as in the 2X but also received a third dose the following morning, 24 h after the first dose was administered and 4 h before the end of the study. Fold changes in protein levels were quantitated by using Odyssey software and were normalized to the levels of the actin loading control.

Fig. 5.

PCI-24781 inhibits homologous recombinational repair. AA8 cells containing the DR-GFP recombination substrate (DRAA8/CHO) were used to determine the effect of PCI-24781 on homologous recombinational repair activity. Repair of I-SceI-induced DSB is quantified by green fluorescence. PCI-24781 was added at the indicated concentrations. The numbers shown in the upper right-hand corners of the graphs are the percentages of HR-positive cells (gated green cells) relative to the total live cell population (red and green). The Western blot (WB) shows levels of the HR protein RAD51 after treatment with the indicated doses of PCI-24781.

Fig. 6.

Pretreatment with PCI-24781 enhances sensitivity to radiation in three tumor cell lines. Colony-formation assays were performed by pretreating cells with 1.0 μM PCI-24781 for various lengths of time, followed by exposure to the indicated doses of γ-irradiation (IR) in HCT116 (A), NCI-H460 (B), and A549 (C) cells. The percentage of survival is represented logarithmically on the ordinate.

Fig. 7.

NHEJ-defective cells are hypersensitive to the effects of PCI-24781. Clonogenic survival for CHO-K1 and its Ku86 mutant derivative after 0.5, 1.0, and 2.0 μM PCI-24781 treatment was assayed. Cells (10³) were plated in 10-cm dishes, allowed to adhere, treated with the indicated doses of PCI-24781 for 24 h, and replaced with drug-free medium.