Original Article | Published:

Genetic cooperation between p21Cip1 and INK4 inhibitors in cellular senescence and tumor suppression

Oncogene volume 26, pages 76657674 (06 December 2007) | Download Citation



Cell-cycle inhibitors of the Cip/Kip and INK4 families are involved in cellular senescence and tumor suppression. Some of these proteins, p21Cip1, p16INK4a and p15INK4b, are coexpressed in response to antiproliferative signals such as cellular senescence resulting in cell-cycle arrest. To understand the roles of these inhibitors and their synergistic effect, we have characterized the growth properties and senescent behavior of primary cells deficient in p21Cip1 and expressing an endogenous Cdk4R24C (cyclin-dependent kinase) mutant (Cdk4R24C knock-in cells) insensitive to INK4 proteins. Inactivation of both p21Cip1 and INK4 pathways strongly cooperate in suppressing cellular senescence in vitro. These double mutant cells behavior as immortal cultures and display high sensitivity to cellular transformation by oncogenes. Moreover, mice double mutant in the INK4 and p21Cip1 pathways (Cdk4R24C; p21Cip1-null mice) display an increased incidence of specific sarcomas, suggesting a significant cooperation between these two families of cell-cycle inhibitors in senescence responses and tumor suppression in vivo.


The involvement of cell-cycle regulators in human cancer has been extensively established in the past years (Sherr, 2000; Malumbres and Carnero, 2003). Most alterations in the cell-cycle target regulators of the G1/S transition, a period where cells decide whether to enter into the cell cycle upon mitogenic stimuli, or staying quiescent in response to antimitogenic or senescence signals. The retinoblastoma protein (pRb) pathway seems to play a key role in the regulation of these cellular processes since pRb and their regulators – cyclins, cyclin-dependent kinases (Cdks) and Cdk inhibitors – are frequently mutated in human cancer (Malumbres and Barbacid, 2001).

In normal cells, pRb proteins repress the transcription of genes required for DNA replication or mitosis and maintain cells in a quiescent state. This function is achieved through the sequestering of inactive E2F transcription factors and through the binding to histone deacetylases and chromatin remodeling complexes (Attwooll et al., 2004; Korenjak and Brehm, 2005; Macaluso et al., 2006). Upon mitogenic stimuli, D-type cyclins are induced and activate the cell-cycle kinases Cdk4 and Cdk6. Cyclin D-Cdk4/6 complexes phosphorylate and partially inactivate pRb, allowing the expression of some E2F target genes such as cyclin E. Induction of cyclin E allows the activation of Cdk2, which is also able to further phosphorylate and completely inactivate pRb, triggering the massive transcription of genes required for DNA replication and mitosis. Cdk2 is also able to bind A-type cyclins during S phase, whereas the control of G2 and M phases mainly depends on cyclin A- and cyclin B-Cdk1 complexes (Sherr, 2000; Malumbres and Barbacid, 2005).

Mitogenic stimuli induce cyclins and therefore activate Cdks, whereas antimitogenic signals arrest this process by inducing members of the two families of Cdk inhibitors (CKIs), the INK4 and Cip/Kip families (Sherr and Roberts, 1999). INK4 proteins (p16INK4a, p15INK4b, p18INK4c and p19INK4d) specifically bind Cdk4 or Cdk6 proteins disturbing their binding to D-type cyclins and forcing a kinase inactive state. Cip/Kip inhibitors (p21Cip1, p27Kip1 and p57Kip2), on the other hand, are able to bind Cdk-cyclin complexes forming ternary structures. CKIs are induced in response to various antimitogenic stimuli. For instance, p21Cip1 has been shown to play a major role in inducing p53-dependent G1 cell-cycle arrest following DNA damage, and it is also induced by transforming growth factor-β, along with p15INK4b, resulting in Cdk inactivation (Malumbres and Carnero, 2003). Specific CKIs are also involved in senescence responses and the ageing-dependent control of proliferation in stem cells and specific progenitors (Janzen et al., 2006; Krishnamurthy et al., 2006; Molofsky et al., 2006). In culture, primary somatic cells undergo a finite number of divisions before they arrest as senescence cells (Hayflick, 1965). This cellular senescence is accompanied by high levels of Cdk inhibitors including p21Cip1, p16INK4a and p15INK4b (Noda et al., 1994; Alcorta et al., 1996; Palmero et al., 1997; Stein et al., 1999). Similarly, these CKIs are induced in the premature senescence phenotype in response to inappropriate oncogenic stimulation. Thus, overexpression of oncogenic Ras in primary cells results in high levels of p16INK4a (Serrano et al., 1997) and p15INK4b (Malumbres et al., 2000), which are accompanied by p19ARF induction and consequent activation of the p53/p21Cip1 pathway (Serrano et al., 1997; Palmero et al., 1998). In fact, overexpression of several INK4 and Cip/Kip proteins is sufficient to arrest cells in G1 and induce a senescent-like phenotype (Serrano et al., 1997; McConnell et al., 1998; Vogt et al., 1998; Malumbres et al., 2000).

Although all these inhibitors participate in the senescence signaling pathways, individual inactivation of specific CKIs do not completely preclude senescence responses. Mice deficient in either p16INK4a, p15INK4b, p18INK4c or p21Cip1 have only a limited susceptibility to tumor development in aged mice (Latres et al., 2000; Krimpenfort et al., 2001; Martin-Caballero et al., 2001; Sharpless et al., 2001). In addition, mouse embryonic fibroblasts (MEFs) without either of these inhibitors arrest in a senescence-like crisis period when cultured in vitro and are mostly resistant to transformation by Ras oncogenes (Pantoja and Serrano, 1999; Latres et al., 2000; Krimpenfort et al., 2001; Sharpless et al., 2001). The ‘weak’ phenotype of p16INK4a and p15INK4b individual mutants could be at least partially explained by compensation between these two inhibitors given the strong functional similarity between them. A double p16INK4a-p15INK4b mutant has not been reported yet since the corresponding genes lie only 12 kb apart in the mouse genome. However, the cooperative effect between these mutations has been investigated using a Cdk4 mutant (Cdk4R24C) protein that is not inhibited by the INK4 proteins (Wolfel et al., 1995). Knock-in cells carrying this Cdk4R24C mutant display proliferative advantages in vitro although they still display significant senescence-like features and resistance to cellular transformation (Sotillo et al., 2001a; Rane et al., 2002). Furthermore, genetic ablation of p21Cip1 in the mouse results in decreased DNA damage responses but not alter immortalization or susceptibility to Ras-induced transformation of primary fibroblasts (Brugarolas et al., 1995; Pantoja and Serrano, 1999). Similarly, disruption of p21Cip1 in diploid human fibroblasts by two sequential rounds of targeted homologous recombination bypass G1 arrest allowing extended lifespan, but not immortalize targeted cells (Brown et al., 1997).

To investigate the relative roles of INK4 and Cip1 inhibitors in cellular senescence and malignant transformation, we have analysed the synergistic effect of combining the insensitivity to INK4 proteins and p21Cip1 deficiency. Overexpression of Cdk4R24C, but not other cell-cycle regulators, efficiently rescues p21Cip1-null cultured cells from senescence-like arrest. Genetic combination of these two alterations in Cdk4R24C knock-in, p21Cip1-null (Cdk4R/R; p21−/−) cells results in cellular immortality and high susceptibility to transformation by Ras oncogenes. Finally, Cdk4R/R; p21−/− mice display increased tumor susceptibility than single mutants, suggesting a significant cooperative effect of deregulating both p21Cip1 and INK4 pathways in tumor progression.


Overexpression of Cdk4R24C rescues p21Cip1-null cells from senescence-like arrest

To identify molecules that cooperate with p21Cip1 in triggering antiproliferative responses, we first overexpressed several wild-type and mutant cell-cycle regulators in p21Cip1-deficient (p21−/−) MEFs at passage 5, immediately preceding the senescence-like state of primary MEFs in culture. As expected, wild-type or p21-null MEFs are not efficient in bypassing the crisis period (Figure 1a). p21−/− cells show a slight advantage in the colony formation assay (22 colonies versus 2 colonies in wild-type cells). However, this proliferative advantage is clearly limited as cells expressing a dominant negative p53 mutant (p53 R175H) form 200 colonies in wild-type cells and more than 500 colonies in p21−/− cells. Among the other molecules tested, only the Cdk4 R24C (Cdk4R24C) mutant displayed a significant activity in this assay. In fact, expression of Cdk4R24C confers a proliferative advantage to p21Cip1-null, but not wild-type cells, comparable to the effect of dominant negative forms of p53 (Figures 1a and b). Overexpression of other cell-cycle regulators such as cyclin D1, cyclin A1, cyclin E1, E2F1, E2F4 or Cdc25A did not rescue proliferation in wild-type or p21Cip1 mutant cells (Figure 1a). These results indicate that enforced expression of Cdk4R24C is not sufficient to confer resistance to senescence-like arrest of wild-type cells, but efficiently cooperates with p21Cip1 deficiency to bypass this crisis period in cultured MEFs.

Figure 1
Figure 1

Cdk4R24C rescues growth arrest p21Cip1-null MEFs. (a) Effect of the indicated cell-cycle regulators on the number of colonies formed in wild-type or p21Cip1-null early-passage primary MEFs. The average±s.d. after three separate experiments is shown. (b) Effect of Cdk4R24C (open triangle), p53R175H (filled squares) or the empty vector (open circles) on cell proliferation in wild-type or p21Cip1-null cells. (c) Cell proliferation of wild-type or p21Cip1-null MEFs after infection with the indicated molecules. The standard 3T3 passage protocol was followed as indicated in Materials and methods. (d) Cdk4R24C expression is continuously required for proliferation of p21Cip1-null cells that have been immortalized in the presence of this INK4-insensitive mutant protein. Cre-mediated deletion of the loxP-conditional Cdk4R24C cDNA results in no colonies, a defect rescued by overexpressing the p53R175H dominant negative mutant. Cdk, cyclin-dependent kinase; MEFs, mouse embryonic fibroblasts.

We then analysed the long-term proliferation rates of pooled cell populations in which we express ectopic Cdk4R24C in primary wild-type or p21−/− MEFs, and compared these results to cells expressing the p53 dominant negative mutant (Figure 1c). Following a 3T3 protocol, wild-type or p21Cip1-null cells do not proliferate beyond passage 6 (about 10–12 population doublings (PD)), although they resume growth approximately at passage 10 (PD 20–24). Ectopic overexpression of Cdk4R24C results in earlier exit from the crisis period, indicating that this mutant contributes to bypass senescence-like arrest but is not sufficient to immortalize cells. In contrast, expression of Cdk4R24C in p21Cip1-null cells completely eliminates the crisis period resulting in a continuous and exponential growth (Figure 1c). Since the only effect of the R24C mutation in Cdk4 is to impede binding of the INK4 inhibitors (Wolfel et al., 1995), these results suggest that elimination of the cell-cycle control by INK4 and p21Cip1 is sufficient to avoid the culture crisis period and to immortalize primary MEFs in culture.

Dependency on Cdk4R24C to maintain growth advantages

We next sought to test whether continuous inactivation of INK4 protein function was required for immortalization. For these experiments, we cloned the Cdk4R24C mutant into pMarx vectors (Hannon et al., 1999), a retroviral vector that contains a Cre recombinase target site (loxP site). Upon integration of the retrovirus into the genome, the loxP site is duplicated such that the genes carried by the virus are flanked on either side by loxP sites. Subsequent expression of Cre recombinase causes excision at these loxP sites, leading to removal from the genome and eventual loss of expression of the exogenous gene (Carnero et al., 2000). Presenescent p21Cip1-null MEFs were infected with the Cdk4R24C mutant in pMarx and immortal cell lines, R24CCR, were generated using a 3T3 protocol. At PD 32, R24CCR cells were infected with a Cre-expressing virus to ablate Cdk4R24C expression. Cells in which INK4 function had been restored failed to form colonies, while control cells continued to proliferate (Figure 1d). Immortalized cell lines that had been generated following infection with non-excisable Cdk4R24C-expressing viruses did not arrest following Cre recombinase expression. Finally, expression of dominant negative p53 from a non-excisable vector in the R24CCR reversible cell line overcame the arrest induced by excision of the Cdk4R24C mutant construct (Figure 1d), indicating that the recovery of mortality could be bypassed by another immortalizing signal such as p53 inactivation. Together, these data indicate that INK4 proteins are still functional in the Cdk4R24C-expressing cells after many passages and they are able to arrest cells if the overexpressed Cdk4R24C mutant is eliminated.

Genetic cooperation between Cdk4R24C and p21Cip1 deficiency

To characterize the synergistic effect of p21Cip1 and INK4 proteins in vivo, we took advantage of Cdk4 R24C (Cdk4R/R) knock-in mice, which express endogenous levels of the Cdk4R24C mutant (Rane et al., 1999). p21Cip1-deficient mice (Brugarolas et al., 1995) were crossed with Cdk4R/R knock-in mice and primary MEFs were isolated from various genotypes. We next analysed the proliferation rates of wild-type (Cdk4+/+; p21+/+), single mutant (Cdk4R/R; p21+/+ or Cdk4+/+; p21−/−) and double mutant (Cdk4R/R; p21−/−) MEFs. As depicted in Figure 2a, deficiency in p21Cip1 (Cdk4+/+; p21−/− cells) or expression of the Cdk4 R24C mutant (Cdk4R/R; p21+/+ cells) confers a relative proliferative advantage to primary MEFs in the early passages. At this stage, both mutations do not show any synergistic effect since proliferation of double mutant cells is similar to wild-type cells expressing the Cdk4 R24C mutant. However, as cells approach the crisis period, the differences are more evident. At passage 5, wild-type MEFs have entered crisis and have lost their capacity to duplicate the populations in culture (population doubling levels (PDLs)=1.1). p21−/− or Cdk4R/R single mutant MEFs show a partial reduction in proliferative potential (PDL=2.6 and 1.6, respectively). However, double mutant Cdk4R/R; p21−/− cells do not show any reduction in the PD capacity and even increase their proliferative potential during passages 4–7 (Figure 2b). Accordingly, wild-type, Cdk4R/R; p21+/+ and Cdk4+/+; p21−/− cells displayed morphological changes characteristic of senescent cells such as flat morphology and enlarged cellular bodies and nuclei (Figure 2c). However, these senescent-like features were barely visible in the actively proliferating Cdk4R/R; p21−/− population, which do not show detectable crisis period and behavior as immortal cells (Figure 2d).

Figure 2
Figure 2

Genetic cooperation between Cdk4R24C and p21Cip1-deficiency in immortalization. Primary MEFs were isolated from wild-type ((+/+)(+/+); empty squares); Cdk4R24C ((R/R)(+/+); gray circles), p21Cip1-null ((+/+)(−/−); empty circles) and double mutant ((R/R)(−/−); filled circles) embryos and cultured following a standard 3T3 protocol. (a) Cell proliferation of early-passage MEFs with the different genotypes. (b) Population doubling levels (PDLs) of wild-type MEFs during the crisis period. Double mutant cells do not display any decrease in PDLs while single mutants show an intermediate phenotype. (c) Representative pictures of cultures of the different genotypes at passage (P) 5. (d) Immortalization of the different MEFs after the 3T3 protocol. Double mutants display no crisis period while single mutants display intermediate phenotypes. Cdk, cyclin-dependent kinase; MEFs, mouse embryonic fibroblasts.

To obtain some insights into the molecular mechanism behind this cooperation, we analysed several cell-cycle regulators in primary MEFs with specific genetic alterations. p21Cip1-null MEFs do not display major alterations in the levels of cell-cycle regulators such as Cdk2, Cdk4, Cdk1, A-type cyclins or the INK4 inhibitors (Figure 3a). Total protein levels of D-type cyclins were reduced in p21Cip1-null cells as described previously (Cheng et al., 1999). This reduction is not due to decreased transcription (since mRNA levels are not reduced; data not shown) but to protein degradation as reported previously (Bagui et al., 2003). p27Kip1 is also slightly increased in p21-null (and Cdk4R/R; p21−/−) cells (Figure 3a). Kinase activities of Cdk2 and Cdk1 are not significantly altered in single or double mutant MEFs. However, at specific cell passages, Cdk4 kinase activity seems to be slightly decreased in Cdk4+/+; p21−/− cells (Figure 3b), as it has been observed previously (Cheng et al., 1999). Binding of p27Kip1 to Cdk4 is not altered in these mutant primary MEFs. The fact that Cdk4R24C is insensitive to INK4 proteins might therefore relief these putative side effects of p21Cip1 absence on Cdk4 activity.

Figure 3
Figure 3

Cell-cycle regulators in Cdk4R24C and p21Cip1 mutant mice. Protein lysates were obtain from wild-type ((+/+)(+/+)); Cdk4R24C ((R/R)(+/+)), p21Cip1-null ((+/+)(−/−)) and double mutant ((R/R)(−/−)) MEFs at P5. (a) Total protein levels for the indicated cell-cycle regulators. (b) Protein lysates were immunoprecipitated (IP) with antibodies against the indicated kinases. The amount of kinase immunoprecipitated and the binding to specific partners was evaluated by western blotting (WB). Kinase activity (KA) was tested in these immunoprecipitates using pRb (Cdk4) or histone 1 (Cdk2 and Cdk1) as substrates. Cdk, cyclin-dependent kinase; MEFs, mouse embryonic fibroblasts; P5, passage 5; pRb, retinoblastoma protein.

Increased oncogenic susceptibility in the absence of INK4 and Cip1 pathways

Since immortalization is a prerequisite for cell transformation, we next analysed whether the combined alteration of Cdk4 and p21Cip1 in these double mutant MEFs cooperates in oncogenic transformation. Transfection of wild-type primary MEFs with oncogenic Ha-Ras does not result in foci formation, which is only achieved using a combination of Ras and E1A in these cells. As described previously (Sotillo et al., 2001a), Cdk4R/R MEFs display a slightly increased susceptibility to oncogenic transformation by Ras in this assay, similarly to that of Cdk4+/+; p21−/− cells (Figure 4a). Interestingly, Cdk4R/R; p21−/− double mutant MEFs display a dramatic increase in cellular transformation by Ras oncogenes (about 97 foci per assay versus 0 in wild-type cells and 4–9 foci in single mutants; Figure 4a). Similarly, double mutant MEFs also display a dramatic increase in the susceptibility to oncogenic transformation by a combination of Ras and E1A oncogenes (about 150 foci in double mutant MEFs versus 18–28 foci in single mutants and 5 foci in wild-type cells). To discard the possibility that Cdk4R/R; p21−/− double mutant cells are already transformed, we analysed the growth of these and other control cells in soft agar. As indicated in Figures 4b and c, none of the indicated genotypes, including Cdk4R/R; p21−/− cells, formed colonies in soft agar, unless these cells were previously transfected with Ras and E1A oncogenes. Thus indicating that Cdk4R/R; p21−/− cells, although immortal, are not transformed.

Figure 4
Figure 4

Double Cdk4R/R; p21−/− MEFs are highly sensitive to cellular transformation. (a) Focus formation assay by Ras or Ras+E1A oncogenes in early-passage MEF with the different genotypes. (b) Growth of selected clones in soft agar and (c) representative pictures. Only clones transfected with Ras+E1A display significant growth in soft agar indicating that non-transfected Cdk4R/R; p21−/− clones do not show oncogenic properties despite their immortal phenotype. Cdk, cyclin-dependent kinase; MEFs, mouse embryonic fibroblasts.

Cooperation between p21Cip1 and INK4 tumor suppressor pathways in vivo

Crosses between p21−/− and Cdk4R/R single mutant mice resulted in the expected ratio of Cdk4R/R; p21−/− double mutant mice (data not shown). Young Cdk4R/R; p21−/− mice present a phenotype similar to that described for Cdk4R/R; p21+/+ mice, including increased size and hyperplasia of some endocrine tissues such as the pancreatic endocrine islets (Rane et al., 1999). Both Cdk4R/R; p21+/+ and Cdk4+/+; p21−/− single mutant mice display certain susceptibility to tumor development as reported previously (Martin-Caballero et al., 2001; Sotillo et al., 2001a; Rane et al., 2002). Spontaneous tumors develop in these animals with long latency, starting at 9 months (Cdk4+/+; p21−/−) or 12 months (Cdk4R/R; p21+/+) of age (Figure 5a). However, by 8–10 weeks, about 15% (5/33) of double mutants die. In all these cases, these animals displayed very long incisor teeth that prevent them from food uptake (data not shown). This feature was also observed in a Cdk4R/R; p21+/– mouse and was not observed in any of the control animals. Whether this phenotype suggests neurological disorders or increased teeth formation is unknown at this moment. In addition to these few animals, additional 15% of Cdk4R/R; p21−/−double mutant mice die earlier than single mutant mice due to a variety of tumors. As indicated in Figure 5a, the average lifespan of Cdk4R/R; p21−/− mice is of 48.5 weeks (versus 65 weeks in Cdk4+/+; p21−/− and 68 weeks in Cdk4R/R; p21+/+ mice). Macroscopic and histologic examination of pathologies in these animals indicated the presence of a variety of tumors in these mutant mice (Table 1). Most of tumors that develop in Cdk4R/R; p21−/− mice are also present in any of the single mutant mice suggesting that cooperation between these two mutations results in a shorter latency rather than new pathologies. The only exception is the presence of a significant increase in osteogenic tumors (11% incidence) develop in the skull of Cdk4R/R; p21−/− double mutant mice but no the other genotypes (Table 1 and Figure 5). In addition to these osteosarcomas, the most frequent tumors in double mutant mice include angiosarcomas, and endocrine pathologies (neoplasis of the Leydig and pancreatic beta cell, and pituitary tumors), in all cases with similar incidence than in Cdk4R/R; p21+/+ mice (Table 1).

Figure 5
Figure 5

Survival and tumor development in Cdk4R/R; p21−/− mice. (a) Survival of mice with different combinations of Cdk4 and p21Cip1 alleles as indicated (Cdk4+/+; p21+/+ (n=25); Cdk4R/R; p21+/+ (n=12); Cdk4+/+; p21−/− (n=11) and Cdk4R/R; p21−/− (n=33)). Microscopic images of hematoxylin and eosin-stained sections are shown (× 120 with insert at × 4000): (b) cementifying fibroma, (c) osteosarcoma, (d) Leydig cell tumor (testis), (e) lung adenocarcinoma, (f) pancreatic endocrine tumor and (g); pars distalis adenoma (pituitary). A macroscopic image of the osteosarcoma is also included in (c). Cdk, cyclin-dependent kinase.

Table 1: Tumor susceptibility in Cdk4R24C and p21Cip1 mutant micea


Unscheduled proliferation in cancer cells frequently results as a consequence of hyperactivation of cell-cycle Cdks through amplification, mutation or overexpression of cyclins and, more frequently, inactivation of Cdk inhibitors (Malumbres and Barbacid, 2001). In fact, inactivation of the INK4 proteins is a frequent oncogenic alteration in several tumor types including sarcomas, lung, liver, pancreatic and hematopoietic malignancies (Malumbres and Barbacid, 2001). Moreover, p27Kip1 downregulation by increased proteolysis is a common event in tumor cells and correlates with poor prognosis. Downregulation of p21Cip1, on the other hand, frequently results from inactivation of the p53 pathway (Malumbres and Carnero, 2003). In addition, this protein can be downregulated in tumor cells by epigenetic alterations in its promoter (Duan et al., 2005). The functional specificity of these two families of Cdk inhibitors has been analysed by different biochemical and genetic means. INK4 proteins specifically bind to monomeric Cdk4 or Cdk6 kinases (Sherr and Roberts, 1999). Genetic studies have shown that INK4 inhibitors do not arrest cell cycle in Cdk4/Cdk6-deficient cells, suggesting a strict functional correspondence between INK4 proteins and the inhibition of Cdk4 and Cdk6 kinase activities (Malumbres et al., 2004). Cip/Kip proteins bind all Cdk-cyclin complexes although it has been proposed that Cdk2 is the preferred target for inhibition. In addition, p21Cip1 and p27Kip1 do not inhibit Cdk4/6 kinase activity in certain circumstances and seem to participate in the stabilization of Cdk4/6-cyclin D complexes (Sherr and Roberts, 1999). However, additional Cdk-independent functions appear to contribute to tumor suppressor activities of p21Cip1 (Coqueret, 2002, 2003; Gregory et al., 2002). In fact, p21Cip1 is able to efficiently arrest the cell cycle in response to several antimitogenic signals in Cdk2-null cells, indicating that Cdk2 is not required for p21Cip1 tumor suppressor activity (Martin et al., 2005).

The tumor suppressor activity of INK4 and Cip1 proteins has been evaluated in vivo using different genetic models in the mouse. Deletion of p21Cip1 results in altered response to DNA damage responses and increased tumor susceptibility, specifically in mesenchymal and hematopoietic cells, at an advanced age (Brugarolas et al., 1995; Martin-Caballero et al., 2001). Absence of p21Cip1 accelerates the development of some tumors such as breast tumors in Ras-transgenic mice (Adnane et al., 2000; Bearss et al., 2002) and intestinal tumors in Apc-haploinsufficient mice (Yang et al., 2001), but has no effect on breast or lymphoid malignancies induced by Myc (Bearss et al., 2002; Martins and Berns, 2002). In addition, p21Cip1 deficiency has contradictory effects on thymic lymphomas and skin tumors (Wang et al., 1997; Philipp et al., 1999; Topley et al., 1999; Weinberg et al., 1999; De la Cueva et al., 2006). Ablation of individual members of the INK4 family similarly results in limited oncogenic phenotype, mostly restricted to reduced susceptibility to tumor development in old p16INK4a or p18INK4c-null mice (Franklin et al., 1998; Latres et al., 2000; Zindy et al., 2000; Krimpenfort et al., 2001; Sharpless et al., 2001). The cooperation between Cip/Kip proteins and INK4 inhibitors has been evaluated using mice deficient in p18INK4c in a p21Cip1- or p27Kip1-null background (Franklin et al., 2000). Loss of both p18INK4c and p21Cip1 results in specific cooperation in pituitary and lung tumors whereas combined ablation of p18INK4c and p27Kip1 significantly increases endocrine (pituitary, adrenals, thyroid, testes and pancreas) and gut tumors (Franklin et al., 2000). Inactivation of these Cip/Kip proteins has also been evaluated in a p16INK4a/ARF-null background. p27Kip1 deficiency in this background results in acceleration of T-cell lymphomas whereas, in contrast, deficiency in p21Cip1 produce no overt alteration in p16INK4a/ARF-null mice (Martin-Caballero et al., 2004). Since these p16INK4a/ARF-null mice display a partial inactivation of the p53 pathway owing to genetic ablation of p19ARF, these results may reflect cooperation between p27Kip1, but not p21Cip1, and the ARF-p53 pathway.

Combined ablation of all the four members of this family has not been evaluated so far. In fact, the close proximity between p16INK4a and p15INK4b genes has prevented so far genetic crosses between knockout mice for these inhibitors. This limitation has been partially overcome by the use of knock-in mice that express Cdk4R24C mutant proteins that are insensitive to all these INK4 inhibitors (Rane et al., 1999). These mice develop a wide spectrum of tumors with complete penetrance including epithelial, mesenquimal and lymphoid malignancies (Sotillo et al., 2001a, 2001b; Rane et al., 2002). Our report provides genetic demonstration that combination of p21Cip1 and INK4 deficiencies (in the Cdk4R24C background) significantly cooperates in development of mesenchymal tumors, and specifically those of bone origin (Figure 5 and Table 1).

The cooperation between Cip1 and INK4 proteins in mesenchymal cells agrees with the cooperative effect of these two pathways on the proliferative properties of fibroblasts in culture and their senescence responses. Senescence-like arrest of human and mouse fibroblasts in culture is mediated by p21Cip1 and members of the INK4 family such as p16INK4a and p15INK4b (Alcorta et al., 1996; Brown et al., 1997; Palmero et al., 1997; Serrano et al., 1997; McConnell et al., 1998; Stein et al., 1999; Malumbres et al., 2000). However, inactivation of individual Cdk inhibitors only partially relieves the senescence response (Medcalf et al., 1996; Pantoja and Serrano, 1999; Latres et al., 2000). We have shown in this report that expression of the INK4-insensitive Cdk4R24C mutant efficiently rescues p21Cip1-null cells from senescence. In addition, combined inactivation of Cip1 and INK4 pathways by genetic means eliminates the senescence response to culture shock in primary fibroblasts. Cdk4R/R; p21−/− mutant cells behavior as immortal cells although they are not transformed. However, they are highly sensitive to transformation by single oncogenes such as Ras. Of note, pRb/p107/p130 triple knockout MEFs are refractory to transformation by Ras oncogenes (Dannenberg et al., 2000; Sage et al., 2000), suggesting that the cooperative effect of p21Cip1 deficiency in a Cdk4R24C background transcends beyond the pRb pathway. In fact, combined alteration of Cdk4 and p21Cip1 equals p53 deficiency and further inactivation of p53 does not seem to cooperate in these assays (data not shown). Moreover, the cooperation between p21Cip1 deficiency and Cdk4R24C results in increased susceptibility to tumor formation in vivo, at least in specific mesenquimal cells. Given the recent demonstration that senescence function as a tumor suppressor mechanism in vivo (Braig and Schmitt, 2006) and the involvement of INK4 inhibitors in age-dependent proliferation (Janzen et al., 2006; Krishnamurthy et al., 2006; Molofsky et al., 2006), these data indicate that concomitant inactivation of p21Cip1 and INK4 inhibitors might help tumor cells to become insensitive to senescence-induced antiproliferative signals.

Materials and methods

Mice and histological analysis

Cdk4R24C knock-in mice (Rane et al., 1999) and p21Cip1-deficient animals (Brugarolas et al., 1995) have been reported previously. These animals were maintained in a mixed 129/Sv (25%) × CD1 (25%) × C57BL/6J (50%) background. Mice were housed at the pathogen-free animal facility of the Centro Nacional de Investigaciones Oncológicas (Madrid) following the animal care standards of the institution. These animals were observed in a daily basis and sick mice were euthanized humanely in accordance with the Guidelines for Humane End Points for Animals used in biomedical research. Tumor latency has been considered equivalent to lifespan. For histological observation, dissected organs were fixed in 10%-buffered formalin (Sigma-Aldrich, St Louis, MO, USA) and embedded in paraffin wax. Three- or five-micrometer-thick sections were stained with hematoxylin and eosin. Additional immunohistochemical examination of the pathologies observed was performed essentially as described in Sotillo et al. (2001a).

Cell culture

MEFs were prepared from E13.5 embryos using standard protocols (Sotillo et al., 2001a). Head and blood organs were removed, and the torso was minced and dispersed in 0.1% trypsin (20 min at 37°C). Cells were grown for two PD and then frozen. MEFs were subcultured 1:4 upon reaching confluence; each passage was considered to be two PDLs. All cultures were maintained in Dulbecco's modified Eagle's medium (Gibco-BRL, Gaithersburg, MD, USA) supplemented with 2 mM glutamine, 1% penicillin/streptomycin and 10% fetal bovine serum (FBS) or donor calf serum.

Growth properties were analysed the classical 3T3 protocol. Every 3 days, cells were trypsinized, counted and 106 cells were plated per 10-cm plate. The relative number of cells is considered as a measure of the number of cells per passage related to the initial number of cells seeded per plate. DNA content was analysed by flow cytometry (Becton-Dickinson, Franklin Lakes, NJ, USA).

Colony formation assays and retroviral-mediated gene transfer

Wild-type MEFs normally undergo senescence in passages 5–6, after 10–12 PD. When seeded at passage 5 (PD 10) and cultured with FBS for additional 12 days, they fail to form colonies. To analyse the effect of different cell-cycle proteins in bypassing cellular senescence in this assay, we infected early passage presenescent MEFs with retrovirus carrying different cell-cycle regulators. Cells were drug-selected for provirus integration and grown until passage 5 (PD 10), and 105 cells were plated in 10-cm dishes and cultured for 12 days. For focus assays, primary MEFs were transfected with RasG12V and/or E1A-expressing vectors as described previously (Sotillo et al., 2001a). Cells were cultivated over 15 days, with medium changed every 3 days, before being fixed and stained with crystal violet.

For retroviral infections, LinXE cells (5 × 106) were plated in a 10-cm dish, incubated for 24 h and then transfected by calcium phosphate precipitation with 20 μg of the retroviral plasmid (16 h at 37°C). After 48 h, the virus-containing medium was filtered (0.45 μm filter; Millipore, Bedford, MA, USA) and supplemented with 8 μg/ml polybrene (Sigma) and an equal volume of fresh media. Target fibroblasts were plated at 8 × 105 cells per 10-cm dish and incubated overnight. For infection, the culture medium was replaced by the appropriate viral supernatant, and then the culture plates were centrifuged (1 h, 1500 r.p.m.) and incubated at 37°C for 16 h. Cultures were selected where indicated with 75 μg/ml hygromycin (Calbiochem, La Jolla, CA, USA), 400 μg/ml G418 (Sigma), 1 μg/ml blasticin (Sigma) or 2 μg/ml puromycin (Fluka, Hauppauge, NY, USA). For the analysis of growth, cells were infected as described above and, at PD 12, 3 × 103 cells were plated in 2.5-cm dishes. At 2–3 day intervals, cells were fixed and stained with crystal violet. After extensive washing, crystal violet was resolubilized in 10% acetic acid and quantified at 595 nm as a relative measure of cell number.

Immunoprecipitation, western blotting and kinase assays

Cells were washed twice with ice-cold phosphate-buffered saline and lysed in NP-40 lysis buffer (150 mM NaCl, 1% NP-40, 50 mM Tris–HCl pH 8.0, 1 mM phenylmethylsulfonyl fluoride, 1 μg/ml leupeptin, 25 μg/ml aprotinin, 1 mM ethylenediamine tetraacetic acid). After 30 min on ice, samples were vortexed (5 min at 4°C) and cleared by centrifugation. Proteins were separated on sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE), transferred to nitrocellulose membranes (Bio-Rad, Richmond, CA, USA), probed using specific antibody and detected using fluorescent donkey (Rockland, Gilbertsville, PA, USA) or goat (Invitrogen, Carlsbad, CA, USA) anti-rabbit secondary antibodies followed by detection using the Odyssey Infrared Imaging System (Li-Cor Biosciences, Lincoln, NE, USA). After transfer of the protein lysates, we probed nitrocellulose membranes with antibodies against Cdk2, Cdk1, Cdk4, cyclin D2, p16INK4a, p21Cip1 and ERK (Santa Cruz Biotechnology, Santa Cruz, CA, USA), cyclin D1 (LabVision, Fremont, CA, USA), p27Kip1 (Transduction Laboratories, Lexington, KY, USA) and α-tubulin (Sigma). For immunoprecipitation, 500 μg of total proteins were incubated with 2 μg of antibody against Cdk2, Cdk4 or Cdk1 during 4 h at 4°C. Then, bound to protein A-sepharose during 1 h. Protein A-sepharose bound proteins were washed three times in lysis buffer, boiled and run using 12% PAGE. Detection of specific proteins was performed as before. Kinase assays were performed essentially as described previously (Martin et al., 2005). A total of 1 μg of mouse pRb protein fragment (amino acids 769–921; Santa Cruz Biotechnology) or histone H1 (Roche, Mannheim, Germany) were used as substrates.


  1. , , , , , . (2000). Loss of p21WAF1/CIP1 accelerates Ras oncogenesis in a transgenic/knockout mammary cancer model. Oncogene 19: 5338–5347.

  2. , , , , , . (1996). Involvement of the cyclin-dependent kinase inhibitor p16 (INK4a) in replicative senescence of normal human fibroblasts. Proc Natl Acad Sci USA 93: 13742–13747.

  3. , , . (2004). The E2F family: specific functions and overlapping interests. EMBO J 23: 4709–4716.

  4. , , , . (2003). P27Kip1 and p21Cip1 are not required for the formation of active D cyclin-cdk4 complexes. Mol Cell Biol 23: 7285–7290.

  5. , , , , . (2002). Differential effects of p21(WAF1/CIP1) deficiency on MMTV-ras and MMTV-myc mammary tumor properties. Cancer Res 62: 2077–2084.

  6. , . (2006). Oncogene-induced senescence: putting the brakes on tumor development. Cancer Res 66: 2881–2884.

  7. , , . (1997). Bypass of senescence after disruption of p21CIP1/WAF1 gene in normal diploid human fibroblasts. Science 277: 831–834.

  8. , , , , , . (1995). Radiation-induced cell cycle arrest compromised by p21 deficiency. Nature 377: 552–557.

  9. , , , . (2000). Loss-of-function genetics in mammalian cells: the p53 tumor suppressor model. Nucleic Acids Res 28: 2234–2241.

  10. , , , , , et al. (1999). The p21(Cip1) and p27(Kip1) CDK ‘inhibitors’ are essential activators of cyclin D-dependent kinases in murine fibroblasts. EMBO J 18: 1571–1583.

  11. . (2002). Linking cyclins to transcriptional control. Gene 299: 35–55.

  12. . (2003). New roles for p21 and p27 cell-cycle inhibitors: a function for each cell compartment? Trends Cell Biol 13: 65–70.

  13. , , , . (2000). Ablation of the retinoblastoma gene family deregulates G(1) control causing immortalization and increased cell turnover under growth-restricting conditions. Genes Dev 14: 3051–3064.

  14. , , , , , et al. (2006). Tumorigenic activity of p21Waf1/Cip1 in thymic lymphoma. Oncogene 25: 4128–4132.

  15. , , , , . (2005). Gfi1 coordinates epigenetic repression of p21Cip/WAF1 by recruitment of histone lysine methyltransferase G9a and histone deacetylase 1. Mol Cell Biol 25: 10338–10351.

  16. , , , , , et al. (1998). CDK inhibitors p18(INK4c) and p27(Kip1) mediate two separate pathways to collaboratively suppress pituitary tumorigenesis. Genes Dev 12: 2899–2911.

  17. , , , , . (2000). Functional collaboration between different cyclin-dependent kinase inhibitors suppresses tumor growth with distinct tissue specificity. Mol Cell Biol 20: 6147–6158.

  18. , , , , , . (2002). Induction of transcription through the p300 CRD1 motif by p21WAF1/CIP1 is core promoter specific and cyclin dependent kinase independent. Cell Cycle 1: 343–350.

  19. , , , , , et al. (1999). MaRX: an approach to genetics in mammalian cells. Science 283: 1129–1130.

  20. . (1965). The limited in vitro lifetime of human diploid cell strains. Exp Cell Res 37: 614–636.

  21. , , , , , et al. (2006). Stem-cell ageing modified by the cyclin-dependent kinase inhibitor p16INK4a. Nature 443: 421–426.

  22. , . (2005). E2F-Rb complexes regulating transcription of genes important for differentiation and development. Curr Opin Genet Dev 15: 520–527.

  23. , , , , . (2001). Loss of p16Ink4a confers susceptibility to metastatic melanoma in mice. Nature 413: 83–86.

  24. , , , , , et al. (2006). p16INK4a induces an age-dependent decline in islet regenerative potential. Nature 443: 453–457.

  25. , , , , , et al. (2000). Limited overlapping roles of P15(INK4b) and P18(INK4c) cell cycle inhibitors in proliferation and tumorigenesis. EMBO J 19: 3496–3506.

  26. , , . (2006). Rb family proteins as modulators of gene expression and new aspects regarding the interaction with chromatin remodeling enzymes. Oncogene 25: 5263–5267.

  27. , . (2001). To cycle or not to cycle: a critical decision in cancer. Nat Rev Cancer 1: 222–231.

  28. , . (2005). Mammalian cyclin-dependent kinases. Trends Biochem Sci 30: 630–641.

  29. , . (2003). Cell cycle deregulation: a common motif in cancer. Prog Cell Cycle Res 5: 5–18.

  30. , , , , , . (2000). Cellular response to oncogenic ras involves induction of the Cdk4 and Cdk6 inhibitor p15(INK4b). Mol Cell Biol 20: 2915–2925.

  31. , , , , , et al. (2004). Mammalian cells cycle without the D-type cyclin-dependent kinases Cdk4 and Cdk6. Cell 118: 493–504.

  32. , , , , . (2004). Different cooperating effect of p21 or p27 deficiency in combination with INK4a/ARF deletion in mice. Oncogene 23: 8231–8237.

  33. , , , , , et al. (2005). Cdk2 is dispensable for cell cycle inhibition and tumor suppression mediated by p27(Kip1) and p21(Cip1). Cancer Cell 7: 591–598.

  34. , , , . (2001). Tumor susceptibility of p21(Waf1/Cip1)-deficient mice. Cancer Res 61: 6234–6238.

  35. , . (2002). Loss of p27(Kip1) but not p21(Cip1) decreases survival and synergizes with MYC in murine lymphomagenesis. EMBO J 21: 3739–3748.

  36. , , , . (1998). Inhibitors of cyclin-dependent kinases induce features of replicative senescence in early passage human diploid fibroblasts. Curr Biol 8: 351–354.

  37. , , . (1996). Expression of p21 is not required for senescence of human fibroblasts. Cancer Res 56: 4582–4585.

  38. , , , , , et al. (2006). Increasing p16INK4a expression decreases forebrain progenitors and neurogenesis during ageing. Nature 443: 448–452.

  39. , , , , . (1994). Cloning of senescent cell-derived inhibitors of DNA synthesis using an expression screen. Exp Cell Res 211: 90–98.

  40. , , , , , et al. (1997). Accumulation of p16INK4a in mouse fibroblasts as a function of replicative senescence and not of retinoblastoma gene status. Oncogene 15: 495–503.

  41. , , . (1998). p19ARF links the tumour suppressor p53 to Ras. Nature 395: 125–126.

  42. , . (1999). Murine fibroblasts lacking p21 undergo senescence and are resistant to transformation by oncogenic Ras. Oncogene 18: 4974–4982.

  43. , , , , . (1999). Tumor suppression by p27Kip1 and p21Cip1 during chemically induced skin carcinogenesis. Oncogene 18: 4689–4698.

  44. , , , . (2002). Germ line transmission of the Cdk4(R24C) mutation facilitates tumorigenesis and escape from cellular senescence. Mol Cell Biol 22: 644–656.

  45. , , , , , et al. (1999). Loss of Cdk4 expression causes insulin-deficient diabetes and Cdk4 activation results in beta-islet cell hyperplasia. Nat Genet 22: 44–52.

  46. , , , , , et al. (2000). Targeted disruption of the three Rb-related genes leads to loss of G(1) control and immortalization. Genes Dev 14: 3037–3050.

  47. , , , , . (1997). Oncogenic ras provokes premature cell senescence associated with accumulation of p53 and p16INK4a. Cell 88: 593–602.

  48. , , , , , et al. (2001). Loss of p16Ink4a with retention of p19Arf predisposes mice to tumorigenesis. Nature 413: 86–91.

  49. . (2000). The Pezcoller lecture: cancer cell cycles revisited. Cancer Res 60: 3689–3695.

  50. , . (1999). CDK inhibitors: positive and negative regulators of G1-phase progression. Genes Dev 13: 1501–1512.

  51. , , , , , et al. (2001a). Wide spectrum of tumors in knock-in mice carrying a Cdk4 protein insensitive to INK4 inhibitors. EMBO J 20: 6637–6647.

  52. , , , , , et al. (2001b). Invasive melanoma in Cdk4-targeted mice. Proc Natl Acad Sci USA 98: 13312–13317.

  53. , , , . (1999). Differential roles for cyclin-dependent kinase inhibitors p21 and p16 in the mechanisms of senescence and differentiation in human fibroblasts. Mol Cell Biol 19: 2109–2117.

  54. , , , , . (1999). p21(WAF1/Cip1) functions as a suppressor of malignant skin tumor formation and a determinant of keratinocyte stem-cell potential. Proc Natl Acad Sci USA 96: 9089–9094.

  55. , , , , . (1998). Independent induction of senescence by p16INK4a and p21CIP1 in spontaneously immortalized human fibroblasts. Cell Growth Differ 9: 139–146.

  56. , , . (1997). Loss of p21 increases sensitivity to ionizing radiation and delays the onset of lymphoma in atm-deficient mice. Proc Natl Acad Sci USA 94: 14590–14595.

  57. , , , , , et al. (1999). Genetic deletion of p21WAF1 enhances papilloma formation but not malignant conversion in experimental mouse skin carcinogenesis. Cancer Res 59: 2050–2054.

  58. , , , , , et al. (1995). A p16INK4a-insensitive CDK4 mutant targeted by cytolytic T lymphocytes in a human melanoma. Science 269: 1281–1284.

  59. , , , , , et al. (2001). Targeted inactivation of the p21(WAF1/cip1) gene enhances Apc-initiated tumor formation and the tumor-promoting activity of a Western-style high-risk diet by altering cell maturation in the intestinal mucosal. Cancer Res 61: 565–569.

  60. , , , , . (2000). INK4d-deficient mice are fertile despite testicular atrophy. Mol Cell Biol 20: 372–378.

Download references


We thank Mariano Barbacid for helpful discussions, and Sheila Rueda and Blanca Velasco for their valuable help in the management of the mouse colony. This work was supported by grants from the INSERM, and Association pour la Recherche contre le Cancer and the Région Aquitaine (to PD); Ministerio de Sanidad (FIS-02/0126), Fundación Mutua Madrileña and the Ministerio de Educación y Ciencia (SAF2005-00944) (to AC); and from the Ministerio de Educación y Ciencia (SAF2006-05186), Fundación Científica de la Asociación Española contra el Cáncer, Fundación Ramón Areces and Fundación Médica Mutua Madrileña Automovilística (to MM). The Cell Division and Cancer Group of the CNIO is supported by the OncoCycle program from the Comunidad de Madrid (S-BIO-0283-2006).

Author information


  1. Cell Division and Cancer Group, Centro Nacional de Investigaciones Oncológicas (CNIO), Madrid, Spain

    • V Quereda
    • , J Martinalbo
    •  & M Malumbres
  2. EA2406, Histologie et Pathologie Moleculaire, University of Bordeaux 2, Bordeaux, France

    • P Dubus
  3. Assays Development Group, CNIO, Madrid, Spain

    • A Carnero


  1. Search for V Quereda in:

  2. Search for J Martinalbo in:

  3. Search for P Dubus in:

  4. Search for A Carnero in:

  5. Search for M Malumbres in:

Corresponding authors

Correspondence to A Carnero or M Malumbres.

About this article

Publication history







Further reading