BMI1, ATM and DDR

The polycomb group protein BMI1 is a component of the polycomb repressive complex 1 (PRC1). Although lacking enzymatic activity, BMI1 makes an important contribution to the intrinsic E3 ubiquitin ligase activity of PRC1 via binding of the catalytic subunit RING2. PRC1 catalyzes mono-ubiquitination of histone H2A at lysine 119 (H2AK119Ub), a well-recognized epigenetic marker associated with gene silencing [1]. A well-studied locus suppressed by BMI1 is ARF/INK4A, which encodes two tumor suppressors p16INK4A and p14/p19ARF with welldemonstrated functions in activating the pRb and p53 pathway, respectively. This suppression thus plays a major role in BMI1-mediated maintenance of the self-renewal of hematopoietic stem cells (HSCs), neural stem cells (NSCs), stem cells of other tissues, and most importantly, cancer stem cells (CSC) [2]. Histone ubiquitination is also an essential aspect of DNA damage response (DDR). In agreement with this knowledge, recent studies reveal a role of BMI1 in DDR regulation. Specifically, through ubiquitination of H2A and γH2AX, BMI1 enhances homologous recombination (HR)-mediated repair of double-stranded breaks (DSBs) [3, 2], an activity that sustains cancer survival during genotoxic agents-based therapies, thereby contributing to BMI1’s ability in promoting tumorigenesis. BMI1 also regulates other aspects of DDR. In general, DDR is constituted of two major intimatelyconnected processes: DNA lesion repair and checkpoint activation. In accordance with this knowledge, we have recently reported a role of BMI1 in compromising DDR checkpoint activation via reducing ATM activation [4]. In response to etoposide-induced DSBs, overexpression of BMI1 decreases γH2AX, ATM S1981 phosphorylation (ATMpS1981), CHK2pT68 (an established ATM target), and G2/M arrest in MCF7 and DU145 cells; knockdown of BMI1 reverses all these events. While the effects on γH2AX suggest that BMI1 indirectly reduces ATMmediated checkpoint activation through enhancing DSB repair, multiple lines of evidence demonstrate otherwise. DSB-elicited ATM activation requires binding of the MRN (MRE11-RAD50-NBS1) complex via a direct association with NBS1 [5]. BMI1 interacts with NBS1; this association alters the binding of NBS1 and ATM [4], which likely attenuates NBS1-mediated ATM activation. This possibility is supported by precipitation of the BMI1NBS1-ATM complex in 293T cells co-transfected with BMI1 and ATM (our unpublished research). Individual deletions of a major nuclear localization site (NLS), HT (helix-turn-helix-turn-helix-turn), PS (proline/serine rich), or RF (ring finger) motif retains BMI1’s ability in binding NBS1 and reducing ATMpS1981 and CHK2pT68 in response to etoposide treatment. However, these mutations affect BMI1’s ability in decreasing γH2AX (Figure 1) [4], a surrogate marker of DSBs. Deletion of NLS (∆NLS) renders BMI1 incapable of downregulating γH2AX in etoposide-treated MCF7 cells, which is likely result of reduced BMI1 nuclear localization. It was found that ∆HT and ∆PS elevated γH2AX (Figure 1), which Editorial