|Home | About | Journals | Submit | Contact Us | Français|
The definition of an oncogene is a gene that actively promotes tumorigenesis. For example, activation of RAS oncogene promotes cell transformation and cancer. Paradoxically, in primary mammalian cells, oncogenic RAS typically triggers cellular senescence, a state of irreversible cell growth arrest. Oncogene-induced senescence is an important tumor suppression mechanism in vivo. Here, we discuss our recent evidence that RAS-induced suppression of DNA repair response via dissociation of BRCA1 from chromatin promotes senescence while predisposing cells to senescence bypass and transformation by allowing for secondary hits. The molecular mechanism we uncovered helps reconcile the tumor-promoting nature of oncogenic RAS with the tumor-suppressing role of oncogene-induced senescence.
Oncogenic mutations in the RAS gene are present in ~30% of human cancers.1 Oncogenic RAS proteins promote cell transformation through the engagement of downstream pathways such as phosphatidylinositol-3 kinase (PI3K)/AKT and the RAF family of serine/threonine kinases such as BRAF.2-4 Paradoxically, activation of the RAS oncogene in primary mammalian cells typically induces a status of irreversible cell growth arrest, known as cellular senescence.5 By driving irreversible growth arrest of cancer progenitor cells harboring the initial oncogenic hit, oncogene-induced senescence is an important tumor suppression mechanism in vivo.5 Further underscoring the importance of senescence in tumor suppression, reactivation of tumor suppressors such as p53 triggers cellular senescence and associated tumor regression due to activation of the innate immune response in mouse models.6 Recently, the senescence-associated secretory phenotype has been proposed to be a cell non-autonomous mechanism by which senescent cells promote transformation of neighboring premalignant cells.5,7 Here, we discuss our recent discovery of a novel cell-intrinsic mechanism by which oncogenic RAS drives cellular senescence while predisposing cells to secondary hits, which ultimately promotes senescence bypass in RAS-expressing primary human cells.8
Downstream effectors of RAS such as BRAF and AKT all possess the ability to induce senescence on their own in primary mammalian cells.9,10 Likewise, inactivation of PTEN, a negative regulator of AKT, also triggers senescence in primary mammalian cells.11 Notably, there are differences in the senescence induced by different oncogenes. For example, it has been demonstrated that the senescence induced by RAS or BRAF is associated with activation of the DNA damage response.12 However, senescence induced by AKT or loss of PTEN is independent of the DNA damage response.13,14 Accordingly, it has been proposed that DNA damage-independent senescence could be utilized as a novel mechanism for developing cancer therapeutics.13,15 Aberrant DNA replication is thought to be the trigger of the DNA damage response during senescence induced by oncogenic RAS mutants.12 Consistently, RAS-induced senescence is dependent upon S phase progression, and inhibition of S phase progression by treating cells with Aphidicolin or contact inhibition blocks RAS-induced senescence.13 In contrast, senescence induced by loss of PTEN is independent of S phase progression and can occur in quiescent cells.13 Thus, the difference in DNA damage response observed in senescence induced by different oncogenes is likely due to their ability to trigger aberrant DNA replication.
Aberrant DNA replication induced by oncogenic RAS is transient and occurs only at the very early stages of senescence.12 However, DNA damage accumulates in fully senescent cells days after aberrant DNA replication in primary human cells, which supposedly have intact DNA repair machinery.12,14 This evidence suggests that the impairment of DNA repair might contribute to the accumulation of DNA damage during RAS-induced senescence. In support of this idea, we discovered that BRCA1 becomes dissociated from chromatin in RAS-infected primary human cells, and BRCA1-mediated DNA repair response is impaired in these cells.8 This observation is unique in that other markers of DNA damage that have been reported in the literature, such as ATM, ATR, Chk1, Chk2, γH2A and 53BP1, are all activated during senescence induced by oncogenic RAS.12,16 Importantly, BRCA1 chromatin dissociation occurs well before the RAS-induced cell cycle exit,8 suggesting that this is not a consequence of the senescence-associated cell cycle exit. Moreover, senescence induced by AKT or loss of PTEN is not associated with BRCA1 chromatin dissociation, correlating with a lack in DNA damage response in these cells.8 This observation further supports the hypothesis that BRCA1 chromatin dissociation is not merely a consequence of senescence. Additionally, BRCA1 chromatin dissociation displays the same kinetics as DNA damage accumulation, and suppression of BRCA1 chromatin dissociation is sufficient to inhibit DNA damage accumulation.8 Further, BRCA1 knockdown triggers DNA damage and senescence in primary human cells.8 Together, these findings further support the premise that BRCA1 chromatin dissociation contributes to the accumulation of DNA damage during RAS-induced senescence.
It is important to note that the levels of activated RAS oncogene determine the outcome of RAS expression. For example, in an inducible H-RASG12V transgenic mouse model, low levels of oncogenic RAS promote cell proliferation, while high levels of oncogenic RAS drive cell senescence.17 Interestingly, tumorigenesis is associated with enrichment of high RAS induction, senescence and senescence bypass.17 Notably, the levels of oncogenic H-RASG12V observed in T24, a human bladder cancer cell line, is sufficient to trigger BRCA1 chromatin dissociation in primary human cells.8 This suggests that physiological level of RAS observed in human cancer cells is sufficient to trigger BRCA1 chromatin dissociation.
As stated previously, oncogene-induced BRCA1 chromatin dissociation well precedes the cell cycle exit during senescence. In addition, oncogenic RAS impairs BRCA1-mediated DNA repair response in cycling RAS-infected cells at a very early stage of senescence.8 Together, a large time window is created for cells to accumulate secondary oncogenic hits prior to the senescence-associated cell cycle exit, which ultimately leads to senescence bypass in a minority of cells while the vast majority of cells eventually exit from the cell cycle and become senescent. Consistent with this model, RAS-expressing cells accumulate significantly greater DNA damage after IR treatment, and IR treatment promotes senescence bypass in RAS-infected cells.8 In mouse models, activated oncogenes (such as RAS or BRAF) have been shown to initially induce senescence and formation of benign lesions. However, these lesions ultimately lead to the development of invasive cancer.17-19 This is consistent with our model whereby oncogenic RAS induces senescence in the vast majority of cells while promoting senescence bypass in a minority of cells, which ultimately outgrow the majority of senescent cells and become transformed.
Interestingly, cells from BRCA1 exon 11 knockout mice display signs of cellular senescence.20 Exon 11 of the mouse BRCA1 gene encodes for the two BRCT repeats of the BRCA1 protein. Recently, it has been demonstrated that BRCT repeats are essential for the tumor suppressive function of BRCA1.21 Notably, we discovered that the C-terminus of BRCA1, which contains two BRCT repeats, is sufficient to be dissociated from chromatin in response to oncogenic RAS.8 BRCT repeats of BRCA1 bind to phosphorylated partners such as BRIP1, Abbarax/RAP80 and CtIP to form three distinct complexes in a mutually exclusive manner.22 These distinct complexes regulate activation of checkpoints in response to DNA damage.22 We found that BRIP1, but not RAP80 or CtIP, is downregulated prior to BRCA1 chromatin dissociation. In addition, BRIP1 knockdown dissociates BRCA1 from chromatin, induces the DNA damage response and triggers cellular senescence. Conversely, ectopic BRIP1 suppresses BRCA1 chromatin dissociation, the DNA damage response and cellular senescence induced by oncogenic RAS.8 Together, these findings suggest a key role of BRIP1 in regulating RAS-induced senescence.
BRIP1 was first described as a member of the DEAH helicase family that binds to the BRCT repeats of BRCA1.23 The interaction between BRIP1 and BRCA1 depends upon the phosphorylation of BRIP1 at residue serine 990, which is regulated in a cell cycle-dependent manner.24 The BRIP1-containing BRCA1 complex is implicated in regulating activation of S phase checkpoint.25 For example, it has been demonstrated that BRIP1 is required for timely S phase progression.25 In addition, BRIP1 and BRCA1 facilitate DNA replication during the S phase of the cell cycle by mediating the loading of CDC45L, the replication-licensing factor.26 Thus, RAS-induced suppression of BRIP1 may promote senescence by driving aberrant DNA replication and activating S phase checkpoints during RAS-induced senescence.
The interaction between BRIP1 and BRCA1 suggests that BRIP1 might be linked to increased cancer risk. Indeed, it has been shown that germline mutations that alter its helicase function or disrupt its interaction with BRCA1 are associated with early-onset breast cancer.27,28 Counter intuitively, the levels of BRIP1 are elevated in breast carcinoma, and higher levels of BRIP1 are associated with higher tumor grade.29 In addition, high levels of BRIP1 positively correlate with HER2 status and expression of Ki67, a marker of cell proliferation.29 It has been demonstrated that HER2 overexpression drives senescence in primary human mammary epithelial cells.30 Thus, overexpression of BRIP1 may contribute to breast cancer by promoting proliferation of mammary epithelial cells through suppressing senescence induced by HER2 overexpression.
As discussed above, IR-treatment in RAS-infected cells promotes senescence bypass.8 In IR-treatment-induced senescence bypassed cells, we show that BRIP1 remains downregulated and BRCA1 remains dissociated from chromatin.8 This is likely due to selection pressure imposed by exogenous DNA damage treatment. In addition, we observed the accumulation of DNA damage in these senescence-bypassed cells compared with controls.8 This suggests that DNA repair remains impaired in these RAS-expressing senescence-bypassed cells induced by IR-treatment. Consistently, it has previously been demonstrated that the DNA damage response is attenuated in RAS-expressing transformed cells.31
Interestingly, we observed rare colonies that bypassed RAS-induced senescence after extended culture without exogenous DNA damage treatment.8 The formation of these colonies may reflect the rare senescence-bypassed cells that are associated with endogenous DNA damage triggered by oncogenic RAS. Thus, it will be interesting to examine the levels of BRIP1 expression and chromatin-associated BRCA1 levels in these cells. Because restoration of BRIP1 expression and suppression of BRCA1 chromatin dissociation inhibits RAS-induced senescence,8 it is plausible that the senescence bypass observed in these rare colonies is due to restoration of BRIP1 expression and, consequently, rescuing BRCA1 chromatin dissociation. Our model would predict that these senescence-bypassed cells with restored BRIP1 expression and chromatin-associated BRCA1 levels would be independent of the DNA damage response. Likewise, it will be interesting to determine the levels of DNA damage in HER2 positive breast carcinoma specimens with elevated levels of BRIP1.29
RAS-induced BRIP1 downregulation that triggers BRCA1 chromatin dissociation was due to suppression of B-Myb expression,8 a known regulator of RAS-induced senescence. For example, ectopic B-Myb suppresses senescence of primary mouse embryonic fibroblasts induced by oncogenic RAS.32 Similar to BRIP1, B-Myb expression is required for S phase entry and acts as a crucial factor for DNA replication during S phase.33 Depletion of B-Myb triggers aberrant DNA replication and the DNA damage response.33 Given its role in S phase entry and progression, it is not surprising that gene amplification or overexpression of B-Myb occurs in several types of cancer.34,35
B-Myb directly regulates BRIP1 by binding to the proximal promoter region of the human BRIP1 gene.8 However, it remains to be determined whether B-Myb downregulation is sufficient to induce BRIP1 downregulation and senescence. To address this question, we developed two individual short hairpins RNA to the human B-Myb gene (shB-Myb). The efficacy of B-Myb knockdown by the shB-Mybs was confirmed by qRT-PCR (Fig. 1A). Indeed, BRIP1 expression was downregulated in B-Myb knockdown cells (Fig. 1B). Further, B-Myb knockdown suppressed the proliferation of primary human fibroblasts as determined by decreased colony formation and BrdU incorporation (Fig. 1C-D). Consistent with the idea that growth inhibition induced by B-Myb knockdown is due to senescence, expression of SA-β-gal, a marker of cellular senescence,36 was induced in B-Myb knockdown cells (Fig. 1E-F). Together, these data further support that premise that B-Myb is a key regulator of BRIP1 during senescence.
In summary, a very early stage of RAS-induced cellular senescence involves suppression of BRIP1 expression to dissociate BRCA1 from chromatin and thereby impair the BRCA1-mediated DNA repair. This promotes cellular senescence and senescence-associated accumulation of DNA damage while also allowing for subsequent secondary hits that may predispose rare clones of cells to escape senescence and ultimately contribute to cell transformation. Therefore, this newly discovered cell-intrinsic pathway reconciles the tumor-promoting nature of oncogenic RAS with the tumor-suppressing role of RAS-induced senescence.
R.Z. is an Ovarian Cancer Research Fund (OCRF) Liz Tilberis Scholar. This work was supported in part by an NCI FCCC-UPenn ovarian cancer SPORE (P50 CA083638) pilot project and SPORE career development award (to R.Z.), a DOD ovarian cancer academy award (OC093420 to R.Z.), a NIH/NCI grant (RO1CA160331 to R.Z.), and an OCRF program project (to R.Z.).
Previously published online: www.landesbioscience.com/journals/smallgtpases/article/19884