Studies in cultured cell lines have shown that oncogene-induced senescence is an important guardian against transformation, but the role of senescence in in vivo tumour formation has not been established. Four papers published in Nature have laid this issue to rest, demonstrating that senescence protects against tumour progression in animal models and in human tissues.

Characterized by the stable growth arrest that occurs in response to various cell stressors, including oncogene induction, senescence is known to be regulated by the ARF–p53 and also the INK4A–RB (retinoblastoma protein) signalling pathways. Chrysilis Michaloglou et al. investigated senescence in human naevi (moles), which are benign tumours of melanocytes that frequently carry activating mutations in the oncogene BRAF. Naevi remain in a growth-arrested state for decades, and only rarely progress into malignancy (melanoma). Michaloglou et al. showed that mutant BRAF-expressing naevi also express high levels of senescence markers and do not proliferate. This is because activated BRAF cannot fully transform human melanocytes — additional mutations are required to disable the senescence machinery, allowing melanomas to arise. Not all naevi cells upregulate INK4A, ARF or p53, however, indicating that other, undiscovered senescence-inducing mechanisms exist in human melanocytes.

In another article, Zhenbang Chen et al. showed that senescent cells are also present in early-stage prostate cancers of both humans and mice. In mouse prostate cells, inactivation of the tumour-suppressor PTEN (phosphate and tensin homologue) leads to senescence through a p53-dependent pathway, both in vitro and in vivo. Loss of both PTEN and p53, however, results in the rapid formation of invasive prostate cancer. So p53-mediated senescence induction seems to be a protective mechanism against prostate tumour progression.

One of the limitations of studying senescent cells in tumours has been the limited number of markers available. Using a microarray screen, Manuel Collado et al. identified a small set of genes whose expression level was correlated with senescence induction. Using a combined panel of markers, his group analyzed the distribution of senescent cells in a various activated RAS-induced tumours in mice. They showed that senescent cells uniformly exist in pre-malignant tumours, but not in malignant ones, and conclude that oncogene-induced senescence helps to restrict tumour progression.

Melanie Braig et al. investigated a novel mechanism of senescence induction in response to oncogenic RAS. Transgenic mice that express activated RAS in their haematopoietic cell compartment (Eμ- NRas mice) develop non-lymphoid neoplasia after long periods of time, with most cells undergoing senescence through an RB-mediated pathway. Previous reports of the presence of heterochromatin foci in senescent cells indicated that histone modifications were associated with growth arrest. Since the histone methyltransferase SUV39H1 binds to RB, the authors investigated whether these two proteins function together to regulate senescence. Indeed, Eμ-NRas mice with targeted deletions of Suv39h1 rapidly succumbed to invasive T-cell lymphomas, indicating that this protein is an important inhibitor of lymphoma progression. The authors propose that SUV39H1 and RB might somehow regulate the DNA packaging required for senescence.

Together, these studies show that although senescence induction is a highly heterogenous process, it seems to be a universal barrier against tumour progression. Further studies are required to determine the exact pathways that trigger senescence when oncogenes become activated, and how some cells escape senescence to progress into highly invasive, metastatic tumours.