The p16INK4A and p14 ARF proteins, encoded by the INK4A-ARF locus, are key regulators of cellular senescence, yet the mechanisms triggering their up-regulation are not well understood. Here, we show that the ability of the oncogene BMI1 to repress the INK4A-ARF locus requires its direct association and is dependent on the continued presence of the EZH2-containing PolycombRepressive Complex 2 (PRC2) complex. Significantly, EZH2 is down-regulated in stressed and senescing populations of cells, coinciding with decreased levels of associated H3K27me3, displacement of BMI1, and activation of transcription. These results provide a model for how the INK4A-ARF locus is activated and how Polycombs contribute to cancer. Cellular senescence is an irreversible growth arrest triggered by several types of stress, including DNA damage, telomere shortening, and oncogene activation (Dimri 2005). Recently, its relevance as a bona fide tumor-suppressive mechanism in vivo has been highlighted (for review, see Narita and Lowe 2005). The Polycomb group (PcG) proteins BMI1, CBX7, and CBX8 are capable of delaying the onset of senescence in mouse and human embryonic fibroblasts (MEFs and HEFs) (Jacobs et al. 1999;Gil et al. 2004;Dietrich et al. 2007). This has been shown to correlate with a decrease in the levels of p16 INK4A and, in some cases, p14 ARF (p19 Arf in mice). Both of these proteins are encoded by the INK4A-ARF locus and are tumor suppressors that act upstream of the pRB and p53 pathways, respectively (Lowe and Sherr 2003).The BMI1-containing Polycomb-Repressive Complex 1 (PRC1), of which many variants are thought to exist, also contains the CBX (CBX2, CBX4, CBX6, CBX7, and CBX8), PHC1-3, RNF1-2, and SCML1-2 proteins (Levine et al. 2004). A second complex, PRC2, contains the histone methyltransferase EZH2, which together with EED and SUZ12 trimethylates histone H3 on Lys 27 (H3K27me3) (Cao and Zhang 2004;Pasini et al. 2004b). The ability of PRC1 to bind to chromatin is dependent on PRC2 function, and it has been proposed that this is primarily achieved via binding to the H3K27me3 mark (Rastelli et al. 1993;Hernandez-Munoz et al. 2005).In this study, we address several outstanding questions concerning the regulation of the INK4A-ARF locus by BMI1. We establish that BMI1 together with other PcGs and the associated H3K27me3 mark "blanket" the locus both in vivo and in vitro (tissue culture) in both mouse and human cells. We show that the repression of the locus by BMI1 is dependent on the continued association of the EZH2-containing PRC2 complex and that the levels of EZH2 are down-regulated in stressed and senescent cells. This down-regulation leads to the loss of H3K27me3, displacement of BMI1, and activation of INK4A transcription, resulting in senescence. Taken together, our results provide a model for how the INK4A-ARF locus is regulated in response to multiple cellular signals and how increased expression of the PcGs contributes to cancer. Results and Discussion PcGs and associated H3K27me3 'blanket' the INK4A-ARF locus ...

Two major mechanisms have been causally implicated in the establishment of cellular senescence: the activation of the DNA damage response (DDR) pathway and the formation of senescence-associated heterochromatic foci (SAHF). Here we show that in human fibroblasts resistant to premature p16INK4a induction, SAHF are preferentially formed following oncogene activation but are not detected during replicative cellular senescence or on exposure to a variety of senescence-inducing stimuli. Oncogene-induced SAHF formation depends on DNA replication and ATR (ataxia telangiectasia and Rad3-related). Inactivation of ATM (ataxia telangiectasia mutated) or p53 allows the proliferation of oncogene-expressing cells that retain increased heterochromatin induction. In human cancers, levels of heterochromatin markers are higher than in normal tissues, and are independent of the proliferative index or stage of the tumours. Pharmacological and genetic perturbation of heterochromatin in oncogene-expressing cells increase DDR signalling and lead to apoptosis. In vivo, a histone deacetylase inhibitor (HDACi) causes heterochromatin relaxation, increased DDR, apoptosis and tumour regression. These results indicate that heterochromatin induced by oncogenic stress restrains DDR and suggest that the use of chromatin-modifying drugs in cancer therapies may benefit from the study of chromatin and DDR status of tumours.

SummaryLung squamous cell carcinoma (LSCC) is a devastating malignancy with no effective treatments, due to its complex genomic profile. Therefore, preclinical models mimicking its salient features are urgently needed. Here we describe mouse models bearing various combinations of genetic lesions predominantly found in human LSCC. We show that SOX2 but not FGFR1 overexpression in tracheobronchial basal cells combined with Cdkn2ab and Pten loss results in LSCC closely resembling the human counterpart. Interestingly, Sox2;Pten;Cdkn2ab mice develop LSCC with a more peripheral location when Club or Alveolar type 2 (AT2) cells are targeted. Our model highlights the essential role of SOX2 in commanding the squamous cell fate from different cells of origin and represents an invaluable tool for developing better intervention strategies.

In mouse and human neural progenitor and glioblastoma "stem-like" cells, we identified key targets of the Polycomb-group protein BMI1 by combining ChIP-seq with in vivo RNAi screening. We discovered that Bmi1 is important in the cellular response to the transforming growth factor-β/bone morphogenetic protein (TGF-β/BMP) and endoplasmic reticulum (ER) stress pathways, in part converging on the Atf3 transcriptional repressor. We show that Atf3 is a tumor-suppressor gene inactivated in human glioblastoma multiforme together with Cbx7 and a few other candidates. Acting downstream of the ER stress and BMP pathways, ATF3 binds to cell-type-specific accessible chromatin preloaded with AP1 and participates in the inhibition of critical oncogenic networks. Our data support the feasibility of combining ChIP-seq and RNAi screens in solid tumors and highlight multiple p16(INK4a)/p19(ARF)-independent functions for Bmi1 in development and cancer.