- Top of page
This Hot Topics review, the second in a new Aging Cell series, discusses articles published in the last year that have stimulated new ideas about the tangled relationship of aging to cancer cell biology. The year's highlights include reports on the ability of Mdm2 mutations to diminish risks of cancer in aging mice, on proliferative competition between oncogenic cells and bone marrow stem cells, and on the role of metalloproteinases in overcoming age-associated barriers to tumor invasion. Of particular interest were three articles showing that diminished activity of the tumor-suppressor gene p16/INK4a, while increasing the risk of cancer mortality, can lead to improved function in several varieties of age-sensitive stem cells.
Cancer is a major age-related disease and threat to the longevity of organisms with renewable tissues such as mammals. Although similar in its trajectory to the degenerative diseases of aging, cancer is not in essence a degenerative disease. Rather, malignant tumors arise from hyperproliferative cells – cells that develop derangements in the beneficial processes designed to promote tissue renewal and repair. Cancer is suppressed by the activities of tumor suppressor genes, some of which act to eliminate or permanently arrest the growth of potential cancer cells through the processes of apoptosis or cellular senescence, respectively.
Several years ago, the idea that at least some tumor suppressor mechanisms can be antagonistically pleiotropic and promote aging phenotypes gained molecular support (Tyner et al., 2002; Maier et al., 2004). These reports suggested that activity of the p53 tumor suppressor protein might suppress cancer at the cost of accelerating aging. These articles were rapidly followed by two others describing how, at least in principle, this antagonism might be avoided (Garcia-Cao et al., 2002; Matheu et al., 2004). In the last year, an additional article reinforced this possibility, and moreover provided a potential target for pharmacologically harnessing the tumor suppressor activity of p53 without pro-aging side-effects (Mendrysa et al., 2006). Also in the last year, three articles described the role of the p16/INK4a tumor suppressor in regulating stem cell proliferation (Janzen et al., 2006; Krishnamurthy et al., 2006; Molofsky et al., 2006), and suggested that age-dependent engagement of the circuits that suppress cancer progression in young mice also contribute to age-related changes in immune system maintenance, olfactory neurogenesis, and pancreatic β-cell function. A fifth article, evaluating competition of bone marrow cells with oncogenic precursors, suggested novel additional links between stem cell aging and susceptibility to neoplastic disease (Bilousova et al., 2005), and a sixth article suggested that both the quality of the tumor and quality of host tissue may determine the course of cancer progression in aged organisms (Reed et al., 2007).
By way of background, the findings that p53 can, under some circumstances, cause phenotypes that resemble some aspects of aging provided a platform from which molecular links between aging and cancer could be studied. p53 is a critically important tumor suppressor protein that acts in the nucleus and mitochondria to orchestrate both the apoptotic and senescence responses to potentially oncogenic insults. Seminal work from the Donehower and Scrable laboratories showed that murine p53, if chronically activated, confers exceptional protection against cancer but also accelerates certain degenerative changes associated with aging (Tyner et al., 2002; Maier et al., 2004). These degenerative changes were proposed to result from the high levels of apoptosis or cellular senescence driven by the persistently active p53. In these studies, p53 was rendered persistently active by creating a truncating mutation or overexpressing a short (activating) isoform in the mouse germ line. Shortly thereafter, the Serrano laboratory showed that accelerated aging is not an inevitable consequence of hypervigilant tumor suppression (Garcia-Cao et al., 2002; Matheu et al., 2004). This group created mice carrying extra copies of the entire p53 locus. This manipulation provided near-normal basal levels of p53 activity, but hyper-p53 activity in response to DNA damage. These mice were also remarkably cancer resistant, but did not show signs of accelerated aging. Likewise, mice carrying extra copies of an unmodified INK4a/ARF locus, which encodes two tumor suppressor proteins, p16/INK4a and ARF, were also cancer resistant with normal longevity. ARF facilitates p53 stabilization after it has been activated by DNA damage or other oncogenic insults, whereas p16/INK4a independently arrests cell proliferation in response to stress or certain oncogenic events.
In 2006, an additional study showed that heightened protection against cancer need not accelerate aging in mice. The Perry laboratory created mice carrying a hypomorphic allele of Mdm2, which facilitates p53 degradation after its activation (Mendrysa et al., 2006). Unlike the mice created by Serrano and colleagues, these mice had somewhat elevated basal levels of p53 activity, as determined by increased levels of p53-dependent target gene expression and p53-dependent apoptosis in some tissues. These mice also were cancer resistant without signs of premature aging.
Collectively, all these articles suggest there may be a set-point of p53 activity at which tumor suppression is enhanced without deleterious effects. The Serrano and Perry studies provide reasons for optimism that it might be possible to devise interventions that retard or prevent the development of cancer without accelerating aging phenotypes. The recent Perry's article is especially encouraging in this regard because small molecule inhibitors of Mdm2 already exist. There is a puzzle to these optimistic studies, however. Cancer is a major cause of death in mice, yet none of the cancer-resistant mice showed an increase in longevity. Are there undetected pathologies that are accelerated or exacerbated by the hypervigilant tumor suppression experienced by these mice? Or are mice that die of cancer only a small step away from dying of something else? and – the inevitable and as yet unanswerable big question – how applicable are these studies in mice to humans? Given the growing interest and activity in this area, it is perhaps not overly optimistic to answer: stay tuned!
The potential antagonism between tumor suppressor mechanisms and aging was recently explored from a somewhat different perspective by three articles published simultaneously in 2006. Also working in mice, three groups studied the proliferative potential of stem or progenitor cells in three tissues that are known to show signs of diminished regeneration or repair during aging. Remarkably, all three tissues – the hematopoietic bone marrow (Janzen et al., 2006), the pancreatic islets (Krishnamurthy et al., 2006), and the forebrain subventricular zone (Molofsky et al., 2006) – showed the same age-related change: an increase in expression of the p16/INK4a tumor suppressor protein. Like p53, p16/INK4a is a potent tumor suppressor that halts cell proliferation in response to a variety of conditions and stresses and can also induce cellular senescence.
In the three tissues examined in these studies, the age-dependent rise in p16/INK4a expression was confined to the stem or progenitor cells. This finding suggests that p16/INK4a expression suppresses the development of cancer, at least in part, by suppressing the proliferation of – presumably – damaged or potentially oncogenic stem and progenitor cells. This finding also suggests that p16/INK4a may account, at least in part, for the age-related decline in proliferative capacity in these tissues. Is there any evidence for these ideas? All three articles used the same tool to answer this question: mice genetically engineered to lack one or both copies of the gene encoding p16/INK4a.
p16/INK4a-deficient mice were generated several years ago and are known to be viable and develop normally. However, these mice are moderately cancer prone, although not nearly as cancer prone as p53-deficient mice, and generally die of cancer after about 18–24 months of age. Thus, despite their predisposition to developing cancer, p16/INK4a-deficient mice can be studied into middle age (15–20 months), a time when the declining regenerative capacity of the hematopoietic, pancreatic islet and brain subventricular zone stem/progenitor cell pools is evident in wild-type mice. In all three tissues, p16/INK4a deficiency at least partly rescued the age-dependent decline in stem/progenitor cell proliferative capacity. Moreover, p16/INK4a-deficient bone marrow-derived hematopoietic stem cells were better able to reconstitute an immune system than age-matched wild-type mice. Likewise, when mice were given a toxin that destroys the insulin-producing β-cells in pancreatic islets, p16/INK4a-deficiency protected older mice from the fatal diabetes that results from this treatment. And p16/INK4a-deficiency partly restored the ability of stem cells from the subventricular zone to differentiate in culture and prevented the age-related decline in olfactory neurogenesis, which is known to originate in the subventricular zone. Taken together, these findings suggest that tumor suppression mediated by p16/INK4a also causes an age-related decline in the regenerative or repair capacity of the hematopoietic system, pancreas and brain.
It is not yet known how p16/INK4a suppresses stem or progenitor cell proliferation. As p16/INK4a can induce cellular senescence, one possibility is that this tumor suppressor causes stem or progenitor cells to permanently arrest growth owing to senescence. It is also possible that p16/INK4a reversibly arrests stem/progenitor cell proliferation, or only retards their progression through the cell cycle. Likewise, very little is known about what causes the age-related rise in p16/INK4a expression. Is it induced by age-related oxidative stress? Or is it induced by hormones, such as growth hormone or insulin-like growth factor 1 (IGF-1), which are known to drive aging? Whatever the case, these studies – like the p53 studies discussed above – reveal the antagonism that can exist between tumor suppressor mechanisms and aging. Time will tell whether it will be possible to mitigate the pro-aging activity of p16/INK4a without facilitating the development of cancer.
A recent study from the DeGregori laboratory suggests another reason why suppression of stem or progenitor cell proliferation might fuel aging phenotypes, in this case, ironically, the development of age-related cancer (Bilousova et al., 2005). Working with transplantable bone marrow progenitor cells, this group showed that potentially oncogenic cells (harboring mutant p53 or the Bcr-Abl oncogene) are at a proliferative disadvantage when transplanted with many proliferation-competent normal cells into the bone marrow of mice in which resident hematopoietic cells were destroyed by irradiation. The competitive disadvantage of the potentially oncogenic cells can be greatly mitigated by conditions (genetic manipulations or DNA synthesis inhibitors) that impair the proliferation of the normal stem or progenitor cells in the recipient bone marrow. Thus, replication-impaired stem or progenitor cells might promote the development of cancer by virtue of being poor competitors to potential cancer cells.
With the caveat that this study relied on experimental manipulations in mice that may or may not reflect naturally occurring oncogenic processes, it nonetheless provides a novel and potentially important insight into understanding why cancer is an age-related disease. If a significant factor in cancer development is the ability of normal cells to out-compete premalignant or malignant cells in stem or progenitor cell niches, then the age-related decline in stem or progenitor cells might provide a tissue environment conducive to the outgrowth of cancer cells. Further, if this scenario holds true, any process that curtails the number of normal proliferative cells during aging becomes doubly deleterious. These processes could result in a reversible arrest of cell proliferation, the permanent growth arrest of cellular senescence, or the elimination of cells by apoptosis or other cell death mechanisms. Once implemented, they would not only contribute to tissue degeneration by limiting regeneration and repair, but would also contribute to cancer by limiting the availability of normal cells that might compete with hyperproliferative and potentially cancerous cells for tissue niches. Needless to say, the field awaits more critical testing of these ideas using other cancer models, other tissues, and especially using young and aged animals.
Finally, cancers that arise in the elderly, especially very old individuals, tend to be less aggressive (indolent) compared to similar cancers that arise in middle-aged individuals. At least part of this indolence has been attributed to the generally diminished ability of older individuals to mount a robust angiogenic response to the tumor (Reed & Edelberg, 2004). Angiogenesis is a crucial, if not essential, step in the development of aggressive malignant tumors (Hanahan & Weinberg, 2000). A new study by Reed et al. (2007) describes murine prostate cancer cells that form aggressive highly vascularized tumors equally well in young and old animals. The tumors secreted very high levels of matrix metalloproteinases, enzymes that degrade interstitial extracellular matrices and therefore can create a tissue microenvironment conducive for both tumor and endothelial cell invasion. These findings indicate that even elderly tissues can support aggressive tumor progression, including vigorous angiogenesis, providing the tumor has acquired a suitable combination of phenotypic changes.
Taken together, the above findings also raise an intriguing possibility – that the development of age-related cancer may have at least three phases. During phase one, the accumulation of potentially oncogenic mutations and the aging tissue microenvironment might together favor the emergence of malignant cells. The aged tissue could act in this capacity because its structure and/or function degenerated, or because it lost a critical number of healthy competitor cells. During phase two, the aging tissue might retard the tumor from progressing to more aggressive phenotypes because it is relatively deficient in processes that tumors often hijack for their own nourishment and development. Angiogenesis is one such process whose role is supported by experimental evidence (Reed & Edelberg, 2004), but there may be others. For example, tumor cell migration and/or invasion might be retarded by age-related cross-links in extracellular matrix components. During phase three, the relatively poor conditions of the aged tissue might exert selective pressure on the tumor cells for the emergence of phenotypes that then allow the tumor cells to flourish in the aged host. This scenario raising the intriguing, albeit highly speculative, possibility that tumors of similar origin and grade might have somewhat different characteristics, depending on whether they developed in a young, middle-aged or advanced-aged host.