Histone deacetylase inhibitors decrease NHEJ both by acetylation of repair factors and trapping of PARP1 at DNA double-strand breaks in chromatin

Highlights

• Histone deacetylase inhibitors (HDACi) decrease c-NHEJ.

• p300/CBP inhibition decreases PARP1 acetylation, and increases c-NHEJ.

• PARP1 is significantly enriched at DSBs following HDACi treatment.

• PARP1 trapping by HDACi decreases access of c-NHEJ factors in chromatin.

• Combination of HDACi and PARP inhibitor induces apoptotic cell death.

Abstract

Histone deacetylase inhibitors (HDACi) induce acetylation of histone and non-histone proteins, and modulate the acetylation of proteins involved in DNA double-strand break (DSB) repair. Non-homologous end-joining (NHEJ) is one of the main pathways for repairing DSBs. Decreased NHEJ activity has been reported with HDACi treatment. However, mechanisms through which these effects are regulated in the context of chromatin are unclear. We show that pan-HDACi, trichostatin A (TSA), causes differential acetylation of DNA repair factors Ku70/Ku80 and poly ADP-ribose polymerase-1 (PARP1), and impairs NHEJ. Repair effects are reversed by treatments with p300/CBP inhibitor C646, with significantly decreased acetylation of PARP1. In keeping with these findings, TSA treatment significantly increases PARP1 binding to DSBs in chromatin. Notably, AML patients treated with HDACi entinostat (MS275) in vivo also show increased formation of poly ADP-ribose (PAR) that co-localizes with DSBs. Further, we demonstrate that PARP1 bound to chromatin increases with duration of TSA exposure, resembling PARP “trapping”. Knockdown of PARP1 inhibits trapping and mitigates HDACi effects on NHEJ. Finally, combination of HDACi with potent PARP inhibitor talazoparib (BMN673) shows a dose-dependent increase in PARP “trapping”, which correlates with increased apoptosis. These results provide a mechanism through which HDACi inhibits deacetylation and increases binding of PARP1 to DSBs, leading to decreased NHEJ and cytotoxicity of leukemia cells.

Introduction

Histone deacetylases (HDACs) catalyze the removal of acetyl groups from the amino-terminal lysine residues of histone proteins, while histone acetyltransferases (HATs) promote their addition, leading to local remodeling of chromatin, a critical step for the access of regulatory proteins to DNA [1]. In addition to their activity on histone proteins, HDACs and HATs also target non-histone proteins. Transcription factors, transcription regulators and DNA repair proteins have been shown to be directly or indirectly regulated by HDACs and HAT enzymes [2]. In leukemias, specific oncogenic fusion proteins recruit HDACs to the promoters of their target genes. These events appear to contribute to leukemogenesis [3], suggesting that these diseases are particularly good candidates for treatment with HDAC inhibitors (HDACi) [4]. Indeed, clinical trials of HDACi in hematological malignancies, such as lymphoma [5] and myelodysplastic syndromes [6] are yielding promising results. Recently, U.S. Food and Drug Administration (FDA) approved HDACi panobinostat for treatment of multiple myeloma [7]. Also, Vorinostat and Romidepsin received FDA approval for treatment of cutaneous T-cell lymphoma [8], [9], [10].

Although HDACi treatment results in apoptosis of cancer cells in vivo, the mechanism underlying this effect is still unclear. It has been reported that HDACi induce apoptosis by upregulation of both death receptors and their ligands in leukemia cells [11], [12], and by activation of the mitochondrial/intrinsic apoptotic pathway through selective activation of Bid and induction of reactive oxygen species (ROS) [13]. In addition, we and others have shown that HDACi cause DNA damage that is significantly increased in solid tumors [14] and leukemias [15] compared with normal cells. Increases in γH2AX and ATM phosphorylation, that are early indicators of DNA damage [16], [17], occur rapidly following HDACi administration [15]. Furthermore, treatment with HDACi such as trichostatin A (TSA) leads to a persistence of DNA double strand breaks (DSBs) [14], [15], and sensitize solid tumor and leukemia cells to ionizing radiation [14], [18], [19], [20].

There are two major mechanisms for the repair of DSBs [21], homologous recombination (HR) and the classical DNA-dependent protein kinase (DNA-PKcs)-dependent non-homologous end-joining (c-NHEJ or NHEJ). Recent studies have also identified an alternative and highly error-prone repair pathway, Alt-NHEJ [22], [23], [24] that appears upregulated in cancer and leukemia cells [25], and is implicated as a mechanism for genomic instability [22], [23], [24]. Poly ADP-ribose polymerase-1 (PARP1), that participates in several DNA repair pathways, including single strand break (SSB) repair, is also a key player for initiation of Alt-NHEJ [26].

Several studies have investigated mechanisms by which HDACi modulate DSB repair [27], [28], [29], [30]. Mouse embryonic fibroblasts mutated for NHEJ proteins Ku80 or DNA Ligase IV (LigIV), but not for the HR protein BRCA2, are hypersensitive to TSA [31]. In human prostate cancer cells, HDACi cause hyperacetylation of Ku proteins and decrease the binding of these proteins to DNA [32]. Vorinostat, which is also a pan-HDACi like TSA, downregulates DNA-PKcs in melanoma cells [14]. We previously observed increased co-localization of DNA-PKcs and γH2AX in HDACi-treated cells indicating that NHEJ components localize in vicinity of DSBs after drug treatment [15]. However, acetylation of Ku70 by HDACi decreases overall NHEJ activity [27], [28], [29]. Thus, it is unclear whether repair factors assembled around DSBs in chromatin following HDAC inhibition are able to access DSBs and perform repair. In this study, we investigated whether HDACi impact recruitment of DSB repair proteins to chromatin, particularly those involved in NHEJ.

We show that pan-HDACi TSA treatment of acute leukemia cells results in differential acetylation of Ku70, Ku80, and PARP1, leading to decreased NHEJ activity. We also show that HAT p300/CBP inhibition using the pharmacological inhibitor C646 reverses these effects, specifically affecting PARP1 acetylation. Moreover, PARP1 is significantly enriched in chromatin following HDACi, compared with Ku proteins, in a mechanism similar to PARP1 trapping observed with PARP inhibitors (PARPis). Notably, PARPis in combination with HDACi lead to further increases in PARP trapping that correlates with increased apoptosis. This suggests that HDACi result in both a physical and functional alteration of PARP1 binding at DSBs, potentially preventing access of NHEJ factors, and therefore resulting in decreased NHEJ repair.