Effects of p21 deletion in mouse models of premature aging

Abstract

An approach to investigate the role of cellular senescence in organismal aging has been to abrogate signaling pathways known to induce cellular senescence and to assess the effects in mouse models of premature aging. Recently, we reported the effect of loss of function of p21, a gene implicated in p53-induced cellular senescence, in the background of the Ku80^−/− premature aging mouse (Zhao et al., EMBO Rep 2009). Here, we provide an overview of the effects of p21 deletion in different models of premature aging.

Aging is the complex process where multiple changes in an organism result in a progressive loss of viability and death.² Aging at the cellular level, also known as cellular senescence, can result from increases in DNA damage, dysfunctional telomeres, decreased mitochondrial function, strong mitogenic signals or changes in chromatin structure.³ Cellular senescence has also been proposed to have evolved as a tumor suppression mechanism, as once a cell becomes senescent it is no longer able to divide.⁴ Other characteristics of a senescent cell include resistance to apoptosis and changes in gene expression such as in those involved in cell cycle regulation and in the secretion of proteases, cytokines and growth factors.³

Cellular senescence induced by increased DNA damage has been theorized to contribute to organismal aging by depleting tissues of cells that are able to proliferate.⁵ In support of this theory, both primary cells from old humans and other primates and tissue samples from old primates show increased amounts of DNA damage and increased numbers of senescent cells.^6–11 Cells from individuals with premature aging syndromes resulting from defective genomic maintenance also exhibit a premature accumulation of DNA damage and undergo accelerated senescence in culture.^12–17 Premature aging conditions in humans include Hutchinson-Gilford progeria syndrome (HGPS), Werner’s syndrome and ataxia-telangiectasia.¹⁴ Similarly, mice with premature aging syndromes have reduced lifespan, increased DNA damage, and cells from such mice also senesce prematurely in culture.^13,18–24

Cellular senescence is, in part, controlled by the p53/p21 tumor suppressor pathway.²⁵ p53 transcriptionally upregulates p21 in response to various types of cellular stress including DNA damage.²⁶ p21 (CDKN1A) is a cyclin dependent kinase (CDK) inhibitor that binds and inactivates Cyclin E/CDK2 and other cyclin/cdk complexes resulting in reversible cell cycle arrest or senescence.²⁷ Levels of p21 increase in normal human fibroblasts undergoing senescence, and upregulation or overexpression of p21 alone is able to induce cellular senescence in both normal human fibroblasts and tumor cells independent of p53.^28–30 Similarly, p21 levels are elevated in prematurely senescing fibroblasts from humans or mice with premature aging syndromes.^1,23,31,32 Conversely, loss of p21 function in normal human fibroblasts results in lifespan extension in culture.³³ Although p21 loss of function leads to an abrogated cell cycle arrest, p21^−/− mice are phenotypically normal³⁴ and no aging phenotypes have been reported. p21^−/− mice do not exhibit an increased cancer incidence for at least 7 months.³⁴ However, p21-deficient mice develop spontaneous tumors at an average age of 16 months, whereas wild-type mice are tumor-free beyond 2 years of age.³⁵

Since absence of p53 or p21 abrogates cell cycle arrest and senescence in response to DNA damage, knocking out p53 or p21 in mouse models of DNA damage and premature aging could help to elucidate the role of senescence in aging. In fact, the deletion of p53 improves aging phenotypes and extends the lifespan of Zmpste24^−/− mice, a premature aging mouse model mimicking human HGPS.³⁶ Moreover, loss of or decreased p53 function rescues embryonic lethality and improves growth defects of mouse embryonic fibroblasts (MEFs) in some models of DNA repair deficiency, including Lig4^−/−, Xrcc4^−/− and Brca1^Δ11/Δ11 mice.^37–39 However, DNA repair deficient mice lacking p53 generally die rapidly due to an increased incidence of malignancy, making it difficult to measure the effects of abrogated senescence resulting from loss of p53 in these mouse models. Several studies have evaluated the effects of p21 loss on premature aging.^1,40–45

Telomerase is important for the maintenance of chromosome ends and consists of a reverse transcriptase, an RNA template (TERC), and associated proteins including dyskerin.⁴⁶ Humans with mutations in telomerase leading to reduced telomerase function, such as in dyskeratosis congenital, show features of premature aging including nail dystrophy, graying hair and bone marrow failure.^46,47 After several generations, mice deficient for telomerase exhibit shortened and dysfunctional telomeres, reduced body weight and premature aging phenotypes including shortened lifespan, alopecia, hair graying and reduced stem and progenitor cell maintenance.^24,48–50 MEFs from such mice also exhibit premature senescence in culture.²¹ Deletion of p53 in Terc^−/− mice with dysfunctional telomeres alleviates the premature senescence observed in their MEFs but leads to a decreased lifespan due to tumors.^31,51 However, loss of p21 function is associated with a median lifespan increase of about 5 months without an increased incidence of cancer.⁴⁰ This increase in lifespan correlates with increased body weight and decreased aging phenotypes. In particular, p21 deletion rescues depletion of stem and progenitor cells in the intestine as well as hematopoietic stem cells (HSCs) in aging late-generation Terc^−/− mice.⁴⁰ These findings suggest that p21 function is responsible for at least some component of the premature aging phenotype seen in late-generation Terc^−/− mice.

Ku80 is critical for DNA repair by nonhomologous end joining and important for telomere maintenance.^52,53 The Ku80^−/− mouse exhibits an early onset of aging characteristics including atrophic skin and hair follicles, osteopenia, premature growth plate closure and a shortened lifespan.²² These mice also exhibit a smaller body size.²² MEFs from Ku80^−/− mice exhibit premature senescence in culture.¹⁸ While p53 deficiency rescues premature senescence in Ku80^−/− MEFs, Ku80^−/−p53^−/− mice exhibit early mortality due to increased tumor incidence.⁵⁴ p21 deficiency also rescues premature senescence in Ku80^−/− MEFs, but does not extend lifespan, improve the Ku80^−/− premature aging phenotypes, or increase body size, despite the absence of increased tumorigenesis.¹ These results suggest that p21-mediated senescence is not the major contributor to premature aging in these mice.

ATM (ataxia telangiectasia mutated) is a phosphoinositide 3-kinase family member upstream of p53 important for transducing the DNA damage signal.⁵⁵ ATM is also important for telomere maintenance.⁵⁶ Humans with ataxia telangiectasia exhibit some features of premature aging, small body size and an increased cancer incidence.⁵⁷ Mice lacking ATM show premature aging phenotypes of the skin and hair, shortened lifespan and a higher incidence of lymphoma.^23,58,59 Atm^−/− MEFs also show premature cellular senescence.²³ Deletion of p53 in Atm^−/− mice abrogates early senescence in MEFs, but a high percentage of these mice die prenatally or develop tumors earlier than either single knockout.⁴⁴ Deletion of p21 in Atm^−/− mice also abrogates early senescence in MEFs.^43,44 Atm^−/−p21^−/− mice develop similarly to Atm^−/− mice, are not more susceptible to tumors and exhibit a delayed onset of lymphomas, presumably due to increased lymphoid cell apoptosis.^43–45 Although body size is not increased and gross development is not improved in Atm^−/−p21^−/− mice, these studies did not specifically analyze aging phenotypes.^44,45

The effects of p21 deletion on the Terc^−/−, Ku80^−/− and Atm^−/− mouse models are summarized in Table 1. Unlike most effects of p53 deletion, p21 deletion does not result in a significant increase in tumors, and in some cases, may delay tumor formation. While aging phenotypes are ameliorated and lifespan is increased in the Terc^−/−p21^−/− mouse model, there is no significant improvement of aging phenotypes, and lifespan is shortened in the Ku80^−/−p21^−/− mouse model. One explanation for the differences observed in these premature aging models with p21 deletion may be the extent or source of DNA damage. While MEFs from Ku80^−/− and late generation Terc^−/− mice show similar amounts of telomere damage, such as chromosome fusions and anaphase bridges, Ku80^−/− MEFs show a higher frequency of chromosome breaks and fragments as well as a more severe cellular senescence phenotype in culture.²¹ These differences in DNA damage are also reflected in the severity of the premature aging phenotypes in each mouse. For example, the median lifespan of Terc^−/− mice is around 15 months; while the more severely DNA damaged Ku80^−/− mice have a median lifespan of just over 5 months. The absence of p21 in Terc^−/− mice does not sensitize the damaged cells to apoptosis, as no increase was seen in the intestinal basal crypts and HSCs.⁴⁰ However, the more severely damaged cells in the Ku80^−/−p21^−/− mice may be further depleted by an increased sensitivity to apoptosis, as was shown in the spleen.¹ The aging process reflects an intricate balance of cellular senescence, apoptosis and tumorigenesis. While p21 deletion is generally not associated with increased tumorigenesis in prematurely aging mice, its effects on organismal aging and lifespan appear to vary depending on the level and/or source of DNA damage.

Table 1.

Effects of p21 deletion in different mouse models of premature aging

Mouse model	Median lifespan	Tumor incidence	Premature aging	Body size	Apoptosis (tissue tested)	MEF premature senescence	Refs.
Terc−/−p21−/−	Increased (from 15 to 20 months)	Not affected	Decreased	Increased	Not affected (intestinal basal crypts, HSCs)	Not reported	40
Ku80−/−p21−/−	Decreased (from 5.1 to 2.5 months)	Not affected	Not affected	Not affected	Increased (spleen)	Abrogated	1
Atm−/−p21−/−	Increased* (from 3.3 to 5.2 months)	Delayed (lymphoma)	Not reported	Not affected	Increased (tumor cells)	Abrogated	41–45

^*Tumor-free median survival.