Sci. Aging Knowl. Environ., 17 August 2005
Vol. 2005, Issue 33, p. pe25
[DOI: 10.1126/sageke.2005.33.pe25]


Making Young Tumors Old: A New Weapon Against Cancer?

Julien Sage

The author is at Stanford University School of Medicine, Stanford CA 94305, USA. E-mail: julsage{at}

Key Words: senescence • cancer • tumor • oncogene • p53 • Ras


Cancer incidence increases dramatically with age (1) (see "Dangerous Liaisons"). However, the mechanisms underlying the connection between aging and tumorigenesis are still not fully understood. One explanation for the age dependency of cancer is that cancer has more chances to develop from old or aging cells because these cells have had more time to acquire tumorigenic mutations. In addition, tumor growth is affected by the interactions between tumor cells and host cells: As an organism ages, changes in the tumor environment may allow tumor cells to escape immune surveillance, proliferate, and spread. In particular, senescent stromal cells in proximity to tumor cells may directly promote tumor growth in a non-cell-autonomous manner by providing growth factors and remodeling the extracellular matrix [reviewed in (2); see also "Faustian Bargain," "Tissue-Tampering Turn-On," and "Led Astray"].

Thus, senescence provides a favorable ground for cancer development. However, cellular senescence has also been proposed to act as a cell-intrinsic tumor suppressor program. This thought-provoking idea comes in part from the first description of cellular senescence by Hayflick and Moorhead in 1961 (3). These seminal experiments showed that primary human fibroblasts "aging" in culture over several generations do not proliferate indefinitely but enter a cell cycle arrest in which they are unresponsive to mitogenic signals (see "More Than a Sum of Our Cells"). We now know that human fibroblasts in culture undergo "replicative senescence" and stop dividing because of a shortening of their telomeres that ultimately triggers a DNA damage response (4). In addition, primary mammalian cells in culture undergo premature senescence when exposed to acute stresses, including DNA damage and strong oncogenic signals. This "stress-induced senescence" resembles replicative senescence in that it involves dramatic morphological alterations, expression of senescence-associated {beta}-galactosidase (SA-{beta}-gal) in lysosomes (5), changes in the structure of chromatin (6), and an irreversible cell cycle arrest. Furthermore, the analysis of the genes controlling senescence supports the idea that senescence acts to suppress tumorigenesis. In particular, the Rb and p53 tumor suppressor genes pathways (see figure 1 in Sharpless Perspective) play a critical role in the establishment and the maintenance of the irreversible cell cycle arrest in senescent cells [see (2, 7-9) for excellent recent reviews].

These observations in cultured cells made it clear that cellular senescence has the potential to suppress cancer by inducing permanent cell cycle arrest in cells with a shortening of telomeres or with an increase in signaling from oncogenes. However, although senescent cells expressing SA-{beta}-gal can be detected in vivo (5), until recently there was only scant evidence that senescence could be triggered in vivo by short telomeres (10), DNA damage (11), and oncogenic stress (12). Although Schmitt and colleagues (11) showed that part of the anticancer action of a chemotherapeutic agent in a mouse model of lymphoma occurs by induction of a senescence program (see "Backup Plan: Senescence Fights Tumors when Apoptosis Fails"), whether cancer progression can be halted by senescence in tumors triggered by oncogenic stimuli had not been determined. Four studies published in the 4 August 2005 issue of Nature now close this gap (13-16).

Premalignant Lesions in Mice Display Markers of Senescence

A short report from Collado et al. describes the phenotype of early tumors induced by the RAS oncogene in mice (13). In this study, the authors show that cells in premalignant lung, pancreas, and skin lesions express SA-{beta}-gal as well as a number of other markers of cellular senescence, including the p16INK4a and p19ARF cell cycle inhibitors (see Sharpless Perspective and Dimri Perspective), and also display changes in their chromatin characteristic of senescence. Advanced tumors isolated from older animals have lost expression of these markers of senescence. These data show that a great number of cells in premalignant lesions induced by the RAS oncogene in mice undergo senescence in vivo and also suggest that loss of function of key cell cycle inhibitors in the Rb and p53 pathways is necessary to bypass senescence and to allow tumor progression.

Oncogene-Induced Senescence in Human Nevi

A second report from Michaloglou and colleagues extend these results to human cancer (14). Human nevi (moles) are thought to arise from a single mutant melanocyte. The authors sought to investigate why nevi stop growing once they reach a certain size and rarely progress to form melanomas. In many cases, the initiating mutation for the formation of a nevus is in the B-RAF gene, which codes for the kinase B-RAF, a direct downstream effector of RAS. The authors show that expression of an oncogenic version of B-RAF (B-RAF containing the mutation Val600->Glu600) is sufficient to induce all the hallmarks of senescence in primary melanocytes in culture after a short period of hyperproliferation. Thus, the authors hypothesized that human nevi are formed of senescent, irreversibly arrested melanocytes. Indeed, human nevi display SA-{beta}-gal activity. In conclusion, the authors propose that activation of B-RAF in melanocytes in vivo triggers an initial wave of proliferation that results in the formation of a nevus. After this period, as sustained activation of B-RAF induces cellular stress, a senescence program is launched, explaining why nevi rarely progress to form malignant tumors. Strikingly, the molecules mediating this senescence phenotype in melanocytes do not appear to be members of the Rb and p53 pathways, which suggests that a number of regulators of senescence remain to be discovered.

High concentrations of members of the RAS pathway have long been known to induce senescence, but such high concentrations are rarely found in human tumors, and expression of oncogenic RAS at endogenous levels in mouse fibroblasts does not induce senescence [(17, 18) and references therein]. Therefore, it is surprising to see that levels of expression of oncogenic RAS (Gly12->Asp12) and RAF (Val600->Glu600) similar to endogenous levels are sufficient to trigger senescence in these two studies (13, 14). This observation may indicate that the cellular response to oncogenic stress is cell-type dependent or that it may vary depending on the cellular microenvironment, and it will be interesting in the future to determine the basis for this variability. Nevertheless, these studies show for the first time that senescence may act as a tumor suppressor mechanism in early tumors initiated by an oncogenic activity.

p53-Dependent Senescence Prevents Progression of Prostate Cancer

In a different model, Chen and colleagues examined the functional relation between loss of the PTEN and p53 tumor suppressor genes in prostate cancer (15). PTEN and p53 are frequently mutated in human prostate cancers. In mice, combined deletion of these two tumor suppressor genes in the prostate epithelium using the Cre/lox system leads to the development of lethal prostate tumors. In contrast, loss of PTEN in prostate epithelial cells only results in less aggressive neoplasias, and loss of p53 has no visible phenotype. How does loss of p53 function cooperate with loss of PTEN? The authors show that in mouse primary fibroblasts in culture and in mouse prostate epithelial cells in vivo, loss of PTEN results in activation of AKT (an oncogene downstream of RAS and normally down-regulated by PTEN), increased concentrations of p19ARF (a mediator of senescence in culture and an activator of p53), stabilization of p53, and induction of senescence. In mice with concomitant loss of p53 and PTEN, senescence markers such as SA-{beta}-gal are absent, which indicates that p53-mediated senescence normally restricts the proliferation of PTEN-mutant cells in vivo. Notably, the tumor suppressor function of p53 in this context is not to induce apoptosis but to induce senescence. Finally, the authors show that SA-{beta}-gal staining can be observed specifically in specimens from early-stage human prostate cancers removed from patients but not in wild-type tissue or advanced prostate tumors. Therefore, progression from early, PTEN-deficient prostate lesions to advanced prostate cancer in vivo requires loss of p53 function to alleviate a senescent cell cycle arrest. These experiments also explain why cancer cells with only one copy of PTEN are selected in early prostate tumors, because loss of the two alleles of PTEN probably induces an irreversible senescence in the earlier stages of prostate cancer.

Senescence Slows Oncogene-Induced Lymphoma Development

In a fourth report, Braig and colleagues studied the importance of senescence as an anticancer mechanism in a mouse model of lymphoma in which tumors are initiated by the expression of an oncogenic form of N-Ras (N-Ras Gly12->Asp12) (16). The authors hypothesize that senescence might be responsible for the slow development of lymphomas in these transgenic mice. Indeed, activation of N-Ras in primary splenocytes induces senescence. What are the mediators of this activation of the senescence program? The authors tested two candidate genes, p53 and Suv39h1. Suv39h1 codes for a histone methylase and is a strong candidate to mediate the changes in chromatin structure observed in senescent cells. In addition, the Suv39h1 methylase is thought to be a key mediator of chromatin changes associated with senescence downstream of the Rb tumor suppressor gene (19, 20). Loss of Suv39h1 function in splenocytes expressing N-Ras Gly12->Asp12 was sufficient to inhibit the trimethylation of lysine 9 of histone 3 (H3K9, one of the modifications of histones controlled by Suv39h1) and the appearance of senescence markers such as SA-{beta}-gal. Mice that express the oncogenic form of N-Ras and lack either p53 or Suv39h1 develop lymphomas much faster than mice that only express oncogenic N-Ras. Further experiments in these double-mutant mice showed that loss of Suv39h1 function does not protect the aggressive lymphomas developing in these mice from cell death induced by a DNA-damaging agent. In contrast, loss of p53 function prevents activation of both the senescence and the apoptosis programs. Together, these experiments demonstrate that senescence is a barrier to lymphoma development in mice. In addition, changes in chromatin structure normally associated with senescence are critical to prevent mutant cells from becoming cancerous.

Together, these four studies show for the first time that endogenous oncogenic signals that are thought to initiate cancer do provoke stress-induced senescence in vivo [more specifically referred to as oncogene-induced senescence (21)]. A model of this process is shown in Fig. 1. In addition, these experiments indicate that, similar to apoptosis, induction of cellular senescence is a strategy employed by mammalian cells to suppress cancer (8).

View larger version (11K):
[in this window]
[in a new window]
Fig. 1. Model for oncogene-induced senescence as a tumor suppressor mechanism. In this model, an oncogenic mutation--such as in the RAS or RAF genes--initiates a first wave of proliferation leading to the formation of a premalignant lesion. Sustained activation of the mutant oncogene triggers premature cellular senescence accompanied by permanent cell cycle arrest. The senescent premalignant lesion may be stable and persist in the organism for a long period of time. Thus, senescence might act as a potent tumor suppressor mechanism in newly formed tumors. The number of cells in a senescent lesion and the time it may take to activate the senescence program might depend on the cell type and the oncogenic mutation. Tumor progression after the first oncogenic hit might require the rapid acquisition of additional mutations in genes controlling the establishment of senescence such as p53 and Rb. It is not clear yet if senescence in the premalignant lesions in vivo is permanent or if it can be reversed by mutations in genes important for the maintenance of the senescence program.

Senescence: A Potential Tool in the Clinic?

Induction of cell death by drugs that cause DNA damage or mitotic catastrophe is a commonly used anticancer strategy in the clinic (8). Do these recent findings mean that clinicians can now also add senescence as a new anticancer weapon, as has been done with cell death? Maybe. A large number of human tumors are deficient for Rb and p53 function, the two major regulators of senescence, indicating that induction of senescence may be impossible to achieve in these tumors. However, two of the studies discussed here offer hope. Michaloglou and colleagues showed that expression of the cell cycle inhibitor p16INK4a does not correlate with cellular senescence in human nevi and that reduction of p16INK4a concentration by RNA interference does not prevent primary melanocytes that express an oncogenic form of B-RAF from undergoing senescence (14). These data suggest that the Rb pathway is not involved in the establishment of stress-induced senescence in melanocytes. In addition, further experiments indicate that the p53 pathway is not induced by activation of B-RAF in senescent melanocytes (14). Therefore, the mediators of the senescent phenotype downstream of B-RAF activation in nevi are still unknown. Once identified, these inducers of senescence may become novel targets for therapies aiming to promote senescence in melanomas and other human tumors. In their mouse lymphoma model, Braig and colleagues show that a key mediator of senescence is the Suv39h1 histone methylase (16). Suvh39h1 is rarely mutated in human cancers; thus, targeted induction of Suv39h1 enzymatic activity could serve as a senescence-inducing therapy in a number of human tumor cells. This observation also suggests that strategies to induce senescence in human tumors should focus on key downstream mediators of the programs controlled by Rb and p53, including chromatin-remodeling enzymes and other cell cycle inhibitors. Interestingly, the Rb family members p107 and p130 may participate in the control of senescence (6, 22) and are not mutated in most human tumors. Thus, induction of p107 and p130 in Rb-deficient human tumors may promote cell cycle arrest and senescence. In conclusion, cellular senescence may act as a key tumor suppressor mechanism in early cancer lesions initiated from mutant cells with activated oncogenes. However, further experiments are required before senescence can be used as a tool in the clinic to fight cancer. The goal of these experiments will be to define more precisely what cellular senescence is, to identify novel markers of senescence, and to characterize the effectors of the senescence program in vivo in various tumor types.

Genetic Links Between Cancer and Aging

These four studies (13-16) underscore the complex relation between cancer and aging, a relation that is growing stronger as genes that were mostly known as cancer genes are now re-discovered to be genes playing a central role in organismal longevity. For instance, loss of p53 function bypasses senescence induced by oncogenes in nascent tumors to promote cancer (15, 16); on the other hand, loss of p53 function rescues premature aging in mutant mice (23), and an isoform of p53 controls longevity in mice (24) (see Campisi Perspective, "Tumor-Free, But Not in the Clear," and Martin Perspective). Similarly, RAS factors and their downstream effectors are potent oncogenes, but RAS signaling is also a critical mediator of insulin action, and RAS activation may thus be pivotal to control longevity (25) (see Longo Perspective). The development of intricate regulatory networks in multicellular organisms ensures proper homeostasis. Modulation of these networks, once we understand them better, will allow us one day to live a long, cancer-free life.

August 17, 2005
  1. L. A. Gloeckler Ries, M. E. Reichman, D. R. Lewis, B. F. Hankey, B. K. Edwards, Cancer survival and incidence from the Surveillance, Epidemiology, and End Results (SEER) program. Oncologist 8, 541-552 (2003).[Abstract/Free Full Text]
  2. J. Campisi, Senescent cells, tumor suppression, and organismal aging: Good citizens, bad neighbors. Cell 120, 513-522 (2005).[CrossRef][Medline]
  3. L. Hayflick, P. S. Moorhead, The serial cultivation of human diploid cell strains. Exp. Cell Res. 25, 585-621 (1961).
  4. F. d'Adda di Fagagna, P. M. Reaper, L. Clay-Farrace, H. Fiegler, P. Carr, T. Von Zglinicki, G. Saretzki, N. P. Carter, S. P. Jackson, A DNA damage checkpoint response in telomere-initiated senescence. Nature 426, 194-198 (2003).[CrossRef][Medline]
  5. G. P. Dimri, X. Lee, G. Basile, M. Acosta, G. Scott, C. Roskelley, E. E. Medrano, M. Linskens, I. Rubelj, O. Pereira-Smith et al., A biomarker that identifies senescent human cells in culture and in aging skin in vivo. Proc. Natl. Acad. Sci. U.S.A. 92, 9363-9367. (1995).[Abstract/Free Full Text]
  6. M. Narita, S. Nunez, E. Heard, M. Narita, A. W. Lin, S. A. Hearn, D. L. Spector, G. J. Hannon, S. W. Lowe, Rb-mediated heterochromatin formation and silencing of E2F target genes during cellular senescence. Cell 113, 703-716 (2003).[CrossRef][Medline]
  7. I. Ben-Porath, R. A. Weinberg, When cells get stressed: An integrative view of cellular senescence. J. Clin. Invest. 113, 8-13 (2004).[CrossRef][Medline]
  8. S. W. Lowe, E. Cepero, G. Evan, Intrinsic tumour suppression. Nature 432, 307-315 (2004).[CrossRef][Medline]
  9. G. P. Dimri, What has senescence got to do with cancer? Cancer Cell 7, 505-512 (2005).[CrossRef][Medline]
  10. A. Satyanarayana, R. A. Greenberg, S. Schaetzlein, J. Buer, K. Masutomi, W. C. Hahn, S. Zimmermann, U. Martens, M. P. Manns, K. L. Rudolph, Mitogen stimulation cooperates with telomere shortening to activate DNA damage responses and senescence signaling. Mol. Cell. Biol. 24, 5459-5474 (2004).[Abstract/Free Full Text]
  11. C. A. Schmitt, J. S. Fridman, M. Yang, S. Lee, E. Baranov, R. M. Hoffman, S. W. Lowe, A Senescence program controlled by p53 and p16(INK4a) contributes to the outcome of cancer therapy. Cell 109, 335-346 (2002).[CrossRef][Medline]
  12. E. Lazzerini Denchi, C. Attwooll, D. Pasini, K. Helin, Deregulated E2F activity induces hyperplasia and senescence-like features in the mouse pituitary gland. Mol. Cell. Biol. 25, 2660-2672 (2005).[Abstract/Free Full Text]
  13. M. Collado, J. Gil, A. Efeyan, C. Guerra, A. J. Schuhmacher, M. Barradas, A. Benguria, A. Zaballos, J. M. Flores, M. Barbacid et al., Tumour biology: Senescence in premalignant tumours. Nature 436, 642 (2005).[CrossRef][Medline]
  14. C. Michaloglou, L. C. Vredeveld, M. S. Soengas, C. Denoyelle, T. Kuilman, C. M. van der Horst, D. M. Majoor, J. W. Shay, W. J. Mooi, D. S. Peeper, BRAFE600-associated senescence-like cell cycle arrest of human naevi. Nature 436, 720-724 (2005).[CrossRef][Medline]
  15. Z. Chen, L. C. Trotman, D. Shaffer, H. K. Lin, Z. A. Dotan, M. Niki, J. A. Koutcher, H. I. Scher, T. Ludwig, W. Gerald et al., Crucial role of p53-dependent cellular senescence in suppression of Pten-deficient tumorigenesis. Nature 436, 725-730 (2005).[CrossRef][Medline]
  16. M. Braig, S. Lee, C. Loddenkemper, C. Rudolph, A. H. Peters, B. Schlegelberger, H. Stein, B. Dorken, T. Jenuwein, C. A. Schmitt, Oncogene-induced senescence as an initial barrier in lymphoma development. Nature 436, 660-665 (2005).[CrossRef][Medline]
  17. D. A. Tuveson, A. T. Shaw, N. A. Willis, D. P. Silver, E. L. Jackson, S. Chang, K. L. Mercer, R. Grochow, H. Hock, D. Crowley et al., Endogenous oncogenic K-ras(G12D) stimulates proliferation and widespread neoplastic and developmental defects. Cancer Cell 5, 375-387 (2004).[CrossRef][Medline]
  18. C. Guerra, N. Mijimolle, A. Dhawahir, P. Dubus, M. Barradas, M. Serrano, V. Campuzano, M. Barbacid, Tumor induction by an endogenous K-ras oncogene is highly dependent on cellular context. Cancer Cell 4, 111-120 (2003).[CrossRef][Medline]
  19. S. J. Nielsen, R. Schneider, U. M. Bauer, A. J. Bannister, A. Morrison, D. O'Carroll, R. Firestein, M. Cleary, T. Jenuwein, R. E. Herrera et al., Rb targets histone H3 methylation and HP1 to promoters. Nature 412, 561-565 (2001).[CrossRef][Medline]
  20. S. Gonzalo, M. Garcia-Cao, M. F. Fraga, G. Schotta, A. H. Peters, S. E. Cotter, R. Eguia, D. C. Dean, M. Esteller, T. Jenuwein et al., Role of the RB1 family in stabilizing histone methylation at constitutive heterochromatin. Nat. Cell Biol. 7, 420-428 (2005).[CrossRef][Medline]
  21. M. Serrano, A. W. Lin, M. E. McCurrach, D. Beach, S. W. Lowe, Oncogenic ras provokes premature cell senescence associated with accumulation of p53 and p16INK4a. Cell 88, 593-602 (1997).[CrossRef][Medline]
  22. J. Sage, G. J. Mulligan, L. D. Attardi, A. Miller, S. Chen, B. Williams, E. Theodorou, T. Jacks, Targeted disruption of the three Rb-related genes leads to loss of G(1) control and immortalization. Genes Dev. 14, 3037-3050 (2000).[Abstract/Free Full Text]
  23. I. Varela, J. Cadinanos, A. M. Pendas, A. Gutierrez-Fernandez, A. R. Folgueras, L. M. Sanchez, Z. Zhou, F. J. Rodriguez, C. L. Stewart, J. A. Vega et al., Accelerated ageing in mice deficient in Zmpste24 protease is linked to p53 signalling activation. Nature, 3 August 2005 [e-pub ahead of print]. doi:10.1038/nature04019
  24. B. Maier, W. Gluba, B. Bernier, T. Turner, K. Mohammad, T. Guise, A. Sutherland, M. Thorner, H. Scrable, Modulation of mammalian life span by the short isoform of p53. Genes Dev. 18, 306-319 (2004).[Abstract/Free Full Text]
  25. V. D. Longo, Ras: The other pro-aging pathway. Sci. Aging Knowledge Environ. 2004 (39), pe36 (2004).[Abstract/Free Full Text]
Citation: J. Sage, Making Young Tumors Old: A New Weapon Against Cancer? Sci. Aging Knowl. Environ. 2005 (33), pe25 (2005).

Cancer Suppression at Old Age.
C. Harding, F. Pompei, E. E. Lee, and R. Wilson (2008)
Cancer Res. 68, 4465-4478
   Abstract »    Full Text »    PDF »

Science of Aging Knowledge Environment. ISSN 1539-6150