Nore1a drives Ras to flick the P53 senescence switch

ABSTRACT

RAS-induced senescence is a protective mechanism to avoid unrestricted cell growth due to aberrant mitogenic signals; however, the exact mechanism by which RAS induces senescence is not known. We recently identified a novel pathway linking RAS to p53 via NORE1A and HIPK2 that mechanistically explains how Ras induces senescence.

Abbreviations

BAX

BCL2-associated X protein

CDKN1A

cyclin-dependent kinase inhibitor 1A (p21^CIP1)

HIPK2

homeodomain interacting protein kinase 2

MEF

mouse embryonic fibroblast

NORE1A

Ras association (RalGDS/AF-6) domain family member 5

p53

tumor protein p53

PI3 kinase

phosphatidylinositol-4,5-bisphosphate 3-kinase

RAS

rat sarcoma viral oncogene homolog

RASSF

Ras association (RalGDS/AF-6) domain

RB

retinoblastoma

Rat sarcoma viral oncogene homolog (RAS) has traditionally been thought of as an oncogene whose activation is a key driving force in the development of multiple tumors. However, somewhat paradoxically, activated Ras can also restrain cell proliferation and induce cell death.¹ Although the growth-promoting functions and pathways of RAS have been well studied and characterized, the exact mechanism(s) by which RAS inhibits proliferation are not as clearly defined and are the subject of much ongoing research. It is now evident, from work by us and others, that in addition to binding to growth promoting effectors such as phosphatidylinositol-4,5-bisphosphate 3-kinase (PI3 kinase) and RAF, RAS also binds to negative effectors that facilitate its prodeath properties.^2,3 Some of these negative RAS effectors are members of the Ras association (RalGDS/AF-6) domain (RASSF) family of tumor suppressors, which all share structural homology.^2,3 These proteins do not have any intrinsic enzymatic activity and are thought to act as scaffolds, integrating multiple proapoptotic signaling pathways, including the Hippo pathway.³ However, the biological role of these RASSF proteins remains only partially understood.

Early work suggested that one way in which RAS exerts its growth inhibitory effects is by inducing senescence, a state of irreversible cell cycle arrest.⁴ The detection of RAS-induced senescence in multiple in vitro and in vivo systems confirms that this phenomenon is physiologically relevant.⁵ It is now well established that senescence provides a barrier that prevents excessive RAS signaling from promoting transformation since malignant tumors almost always lose the senescent phenotype.⁵ However, the exact mechanisms by which RAS promotes senescence and how these mechanisms are bypassed during tumor development remain poorly understood.

One factor that was implicated early on as a critical mediator of RAS-induced senescence is tumor protein p53 (TP53, best known as p53).⁴ While RAS requires intact p53 signaling to mediate senescence and induces post-translational modifications of p53 that are associated with its prosenescence effects,⁶ exactly how RAS communicates with p53 to drive senescence remains unclear. Previous work from our group hinted at a possible link between RAS and p53 involving the Ras association (RalGDS/AF-6) domain family member 5 (RASSF5, also known as NORE1A) tumor suppressor, one of the better characterized RASSF family members. NORE1A binds directly to RAS and promotes apoptosis and cell cycle arrest.^3,7 NORE1A is frequently downregulated in human tumors and inactivation of NORE1A in primary tumors is often associated with an increase in RAS activity. Intriguingly, NORE1A knockout mouse embryo fibroblasts (MEFS) are susceptible to transformation by RAS, which is not the case for wild-type MEFS.⁸ Thus, NORE1A is a bone fide RAS effector that may serve to promote the prodeath properties of RAS while restraining its progrowth/transforming effects. Earlier work from our group showed that NORE1A induces a p53-dependent activation of cyclin-dependent kinase inhibitor 1A (CDKN1A, known as p21^CIP1), a cyclin-dependent kinase inhibitor associated with p53-mediated senescence.⁹ Importantly, downregulation of NORE1A and inactivation of p53 are mutually exclusive, suggesting that they lie in the same pathway.⁹ Thus, NORE1A may be a key factor linking RAS to p53 and a critical mediator of RAS-induced senescence.

Our recent studies have provided further insight into how NORE1A acts to promote RAS-induced senescence.¹⁰ We found that NORE1A is a potent inducer of RAS-mediated senescence, and loss of p53 severely impaired the ability of NORE1A to induce senescence.¹⁰ Mechanistically, RAS promotes the association of NORE1A with homeodomain interacting protein kinase 2 (HIPK2), a kinase that modulates the apoptotic or senescent activity of p53 by promoting specific post-translational modifications of p53. Furthermore, NORE1A scaffolds HIPK2 to p53 in a RAS-dependent manner, inducing acetylation of p53 at lysines 382 and 320 (post-translational modifications of p53 associated with its pro-senescent activity) while simultaneously inhibiting proapoptotic phosphorylation of p53 at serine 46. Importantly, NORE1A expression levels correlated with acetylated p53 in primary human tumors. Consequently, NORE1A induced the expression of p53-regulated prosenescence genes, such as p21^CIP1, and inhibition of p53-regulated proapoptotic genes such as BCL2-associated X protein (BAX).¹⁰

Our data therefore suggest a novel RAS signaling pathway that allows RAS to qualitatively control the post-translational modifications of p53 via a NORE1A/HIPK2 complex, favoring its senescence activity over its apoptotic functions (Fig. 1). This also reveals mechanistically how RAS can promote p53 acetylation, p21^CIP activation, and senescence.⁴ Loss of NORE1A subverts this pathway, uncoupling RAS from p53 and allowing RAS progrowth signals to predominate, thus facilitating transformation. NORE1A is downregulated in many human tumors² and inactivation of NORE1A in hepatocellular carcinomas correlates with reduced p53 acetylation,¹⁰ confirming that this senescence pathway must be disrupted for malignant transformation. NORE1A is a critical component of this new senescence pathway and illustrates why RAS-driven tumors are frequently associated with reduced NORE1A expression.

Figure 1.

Molecular mechanisms whereby RAS promotes cell senescence. In normal cells, RAS promotes the acetylation of p53 via NORE1A and HIPK2, which specifically activates p53 senescence activity (left panel) and serves as a protective mechanism to prevent unrestricted RAS-mediated proliferation. In the absence of NORE1A (right panel), the interaction of HIPK2 with p53 and concomitant p53 senescence-associated acetylation is reduced, resulting in bypass of this senescence signaling pathway and allowing RAS-mediated proliferative signals to predominate. Ac, acetylated lysine; P, phosphorylation.

This may not be the only mechanism by which NORE1A mediates RAS-induced senescence. Intriguingly, we observed that suppression of p53 expression did not completely abrogate the ability of NORE1A to induce senescence,¹⁰ alluding to the existence of additional NORE1A-regulated senescence pathways. The other major senescence effector pathway utilized by RAS involves the retinoblastoma 1 (RB1, known as RB) tumor suppressor,⁴ so it will be interesting to determine whether NORE1A can also mediate RAS-induced senescence via RB.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Funding

Financial support NIH: RR18733, CA133171-01 and NCI intramural funds.