| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Advances in Brief |
Departments of Adult Oncology, Medicine and Genetics, Dana-Farber Cancer Institute and Harvard Medical School [N. E. S., S. A., S. C., D. P. S., D. H. C., R. A. D.], and Department of Pathology, Brigham and Women’s Hospital, Harvard Medical School [D. H. C.], Boston, Massachusetts 02115
 |
ABSTRACT |
|---|
 Top ABSTRACT IntroductionMaterials and MethodsResultsDiscussionREFERENCES |
|---|
|
Introduction |
|---|
|
TopABSTRACTIntroductionMaterials and MethodsResultsDiscussionREFERENCES |
|---|
|
Materials and Methods |
|---|
|
TopABSTRACTIntroductionMaterials and MethodsResultsDiscussionREFERENCES |
|---|
75% FVB/n) were mated with p53+/- females (n >10 FVB/n; Ref. 12
). Nonlittermate p16INK4a+/- p53+/- females were then mated with p16INK4a+/- p53+/- or p16INK4a+/- p53-/- males to generate the experimental colonies (87.5% FVB/n). Littermate controls were analyzed in all instances for tumor-free survival. Animals were genotyped by PCR and monitored as described previously (11
, 12)
. Histological characterization was done as described previously (11)
. We did not perform immunohistochemical analysis on the majority of tumors in this work, and, therefore, many tumors were classified as malignant spindle cell neoplasms. These tumors generally do not express markers of specific mesenchymal differentiation (Myf, S100, desmin, and CD31) yet most likely represent poorly differentiated sarcomas and are termed "fibrosarcoma" or "malignant fibrous histiocytoma" by others. It is possible that a minority of these tumors represent poorly differentiated squamous cell carcinomas, amelanotic melanomas (particularly given that this analysis was performed on albino mice), or other poorly differentiated malignancies.
Cellular Analysis.
MEFs were generated from day 13.5 embryos and grown in DMEM with 10% FCS (Hyclone), 50 µM 2-mercaptoethanol, and Pen/Strep. Growth curves were preformed as described previously (11)
using a VSV-G pseudotype retrovirus concentrated by centrifugation and encoding a dominant-negative mutant of human p53 (V143A). For modified 3T9 analysis, 9 x 105 cells were passaged into either 6-, 10-, and 15-cm dishes every 3 days. Four independent lines were assayed per density. UV treatment and MSP of p16INK4a were performed as described previously (11)
. For hypoxia experiments, 25,000 cells/well in 12-well plates were grown at 1% O2, and cell survival was measured at times indicated. For the high-density experiments, six replicate plates of 1 x 106 cells (p16INK4a-/- and +/+, with and without DNp53) were grown in 10-cm dishes; two lines per genotype were assayed. Cells were refed every other day, and two plates per genotype were harvested on the indicated days. Cells were counted using trypan blue, and cell cycle profile was determined by Propridium Iodide staining. Transformation assays were performed as described previously (11)
. In brief, early passage (passages four to six) littermate cultures (eight lines per genotype in five independent experiments) were transfected with H-Ras(G12V), and either p53-DD (M. Oren) or SV40 Large T-Ag (J. DeCaprio), and then refed but not passaged for 10 days.
Tumor Analysis.
Primary tumors from 18 p53+/+ and 17 p53-/- were analyzed; the distribution of p16INK4a genotypes is shown in Table 2a
. Western blotting for p53, Rb, p16INK4a, and p19ARF was preformed as described on primary tumors (11)
. In brief, tumors were lysed in EBC + protease inhibitors, and cell lysates (50 µg) were resolved on either 16% Tris Glycine or 4–12% NuPage (Novex) gels. In addition to antibodies described previously, membranes were also blotted for Mdm2 (1:200 2A10, gift from A. Levine). LOH analysis of the p16INK4a and p53 loci and MSP were performed as described previously (11)
. For the purpose of Table 2a
, tumors were considered to have an Rb pathway lesion if they lacked expression of p16INK4a or Rb. Tumors were considered to have a p53 lesion if: (a) they lacked detectable p53 and demonstrated increased p19ARF expression (e.g., tumor #9; note p19ARF is repressed by functional p53); (b) overepxressed p53 with low mdm2 (e.g., tumor #6; note mdm2 is a p53-inducible gene); (c) overexpressed mdm2 (e.g., tumor #11); or (d) lacked p19ARF expression with low p53 (e.g., tumor #3).
|
|
Results |
|---|
|
TopABSTRACTIntroductionMaterials and MethodsResultsDiscussionREFERENCES |
|---|
To address further this possibility, replicate cultures of early passage MEFs (p16INK4a-/- and +/+, with and without DNp53) were plated and refed, but not passaged, for 21 days. Under these conditions, plates were confluent by day 3. Plates were harvested, cells were counted, and cell cycle profiles were measured on the indicated days; the ratios of p16INK4a-/-:+/+ cells are shown (Fig. 1a)
. Consistent with a known role for p16INK4a in constraining G1 exit, S phase fraction was higher in p16INK4a-/- than +/+ cells (without DNp53) on days 7 (27.1 versus 19.8%), 11 (8.3 versus 4.3%), and 14 (4.3 versus 2.9%). No difference in cell size (as determined by forward scatter) was discerned between p16INK4a+/+ and -/- cells on any day of culture. This growth difference was of comparable magnitude regardless of DNp53 status (Fig. 1a)
. Additionally, this growth difference is not solely attributable to accumulation of p16INK4a after prolonged culture as p16INK4a-/- MEFs grow with similar kinetics as +/+ cultures up to passage 8 (i.e., 24 days in culture) in a 3T9 assay, where the cells are serially passaged (11)
. Along these lines, no difference in the growth of p16INK4a-/- versus +/+ MEFs was noted under a variety of other settings, including after exposure to UV light or ionizing radiation, under hypoxic conditions (1% O2), in the setting of H-Ras(G12V) expression, and at low seeding density or low serum (data not shown and Refs. 10
and 11
). These data demonstrate that p16INK4a plays a role in constraining cellular proliferation at high density, even in cells lacking functional p53.
|
). In 6-cm dishes, cultures were confluent at all times; in 10-cm dishes, cultures only became confluent by the 3rd day after passaging; in 15-cm dishes, cultures were never confluent. In 10-cm dishes, cells grew considerably better than cells in either 6- or 15-cm dishes. Western blot analysis (Fig. 1c)
of these cultures at passages 3, 5, and 7 showed that p19ARF accumulation increased more rapidly in cultures passaged at lower density, whereas p16INK4a accumulated to slightly higher levels by passage 7 in the highest density cultures. On the other hand, p19ARF expression was markedly reduced in cultures grown at highest density relative to those grown at low density (Fig. 1c)
. Previously, it has been demonstrated that p19ARF (or p14ARF) is induced by E2F (13)
, but p19ARF accumulated to highest levels in the lowest density cultures (Fig. 1c)
. Therefore, the increase in p19ARF levels seen in less dense cultures could not be solely attributed to increased proliferation, as the least dense cultures (15 cm) did not proliferate as extensively as the intermediate density cultures (10 cm; Fig. 1b
). These results suggest that the role of p16INK4a in the regulation of MEF growth only becomes apparent when it is induced to sufficient levels (e.g., by prolonged growth at high density) and primarily in the setting of low p19ARF levels or compromised p53 function.
p16INK4a Loss Augments the Transformation of p53-deficient Cells.
These cooperative effects of p16INK4a and p53 loss in the growth of primary cells prompted an examination of their interactions in modulating foci formation, particularly because focus formation in part reflects unrestricted growth at high density. Classical transformation assays were performed on early passage MEFs by transfecting in combination H-RAS(G12V), a dominant-negative mutant form of p53 (p53-DD), or SV40 large T-Ag (Fig. 1d)
. A modest but consistent 2-fold increase was noted in foci formation in p16INK4a-/- cells transfected with p53-DD and H-Ras (P = 0.025). In contrast, no difference in foci number was noted in T-Ag and H-Ras(G12V) cotransfections; the latter was consistent with the ability of T-Ag to inactivate Rb, thereby rendering p16INK4a status irrelevant. Likewise, p16INK4a loss did not obviate the need for H-Ras(G12V) in this assay, as p16INK4a-/- cells transfected with p53-DD or T-Ag alone did not form foci (data not shown). Transformed foci from H-Ras(G12V) and p53-DD transfections could be subcloned with comparable frequency regardless of p16INK4a genotype (18 of 18 for p16INK4a-/- versus 17 of 18 for p16INK4a+/+). These data demonstrate that the loss of p16INK4a function can enhance the growth and transformation of cells harboring compromised p53 function.
p53 and p16INK4a Loss Cooperate to Increase the Rate of Tumorigenesis and Expand the Spectrum of Tumor Types.
Given the in vitro cooperation in transformation assay between p16INK4a and p53 loss, we sought to extend these observations by assessing the effects of dual p53 and p16INK4a inactivation on tumorigenesis in vivo. Toward this end, cohorts of animals harboring the p16INK4a-/- allele on the p53+/- or -/- backgrounds were monitored (Figs. 2, a–c
and Table 1
). As shown previously, mice lacking either one or both copies of p16INK4a are more tumor prone than littermate wild-type animals, with p16INK4a-/- mice developing sarcomas of various types, lymphomas, and melanomas with a mean latency of 76 weeks (Fig. 2a
and Table 1
). Median tumor latency was shortened in the p53+/- (42 weeks) and -/- (10 weeks) cohorts (Fig. 2, b and c
and Table 1
). Regardless of p53 genotype, p16INK4a-/- mice were more tumor prone than p16INK4a+/+ mice, with heterozygous animals demonstrating an intermediate phenotype. Notably, p16INK4a-/- mice showed a marked increase in the development of angiosarcoma, particularly in the p53-/- background. Of all p16INK4a+/+ mice analyzed (p53+/+, +/-, or -/-, n = 62 mice), there were 2 angiosarcomas (both in p16INK4a+/+ p53-/- animals), as opposed to 15 in p16INK4a-/- mice (n = 77 mice, P = 0.004). Furthermore, in the p16INK4a-/- p53-/- background, angiosarcoma was the most commonly observed tumor (in 56% of mice), whereas p16INK4a+/+ p53-/- developed predominantly thymic lymphoma as described (Table 1
; Refs. 12
and 14
). Angiosarcomas from p16INK4a-/- mice appeared more likely to metastasize in the p53-/- background (5 of 9 p16INK4a-/- tumors versus 3 of 12 from p16INK4a+/+ or +/-; Table 1
), suggesting p16INK4a loss contributes to an aggressive progression-prone phenotype. These experiments were, however, performed in 129 x FVB hybrid animals (see "Materials and Methods"), and, therefore, it is conceivable that the increased susceptibility to angiosarcoma may reflect strain-specific modifiers unequally distributed between the littermate cohorts. This appears to be less likely given the similar findings of an increased incidence of angiosarcoma in p53-/- mice placed on the unrelated BALB/c background (15)
, which harbors a hypomorphic p16INK4a allele (16)
.
|
|
and Table 2a
). With regard to the p16INK4a-Rb pathway, 34 of 34 informative tumors analyzed had detectable Rb (data not shown). Nearly all (9 of 10) of the p16INK4a+/- p53+/+ tumors demonstrated weak or undetectable levels of p16INK4a, frequently with evidence of concomitant p16INK4a promoter methylation (Figs. 2, d and e
, see below). In accord with previous results (17)
, however, the p53-/- background partially obviated the need for p16INK4a loss, as the majority (8 of 15) of tumors from either p16INK4a+/- p53-/- or p16INK4a+/+ p53-/- mice retained high-level expression of p16INK4a. Retained expression of p16INK4a from +/- and +/+ mice could be seen in all three major tumor types of p53-/- mice (four of six angiosarcomas, two of five thymic lymphomas, and two of four malignant spindle cell neoplasms). Sequencing of p16INK4a-overexpressing tumors revealed an absence of point mutations (zero of six). Pooling the present results with our previous analyses (11)
revealed a statistically significant difference in the frequency of p16INK4a expression in tumors from p53+/+ versus p53-/- mice (P = 0.01; Table 2b
). These data imply that either tumors from p53-/- mice do not require Rb pathway lesions or, as suggested previously (17)
, sustain lesions other than Rb and p16INK4a loss per se to circumvent this pathway.
|
, and see "Materials and Methods"), regardless of p16INK4a genotype. Of tumors from p53+/+ mice, the most commonly observed lesions were either elevation of p53 in the setting of undetectable Mdm2 expression (e.g., tumor #7; Fig. 2d
) consistent with p53 point mutation or undetectable p19ARF and p53 expression (e.g., tumor #8) consistent with p19ARF loss (either genetic or epigenetic). Three of three tumors from p53+/- mice demonstr ated reduction to homozygosity (i.e., LOH) of p53 by Southern analysis (data not shown). These genetic data complement epidemiological and biochemical data, indicating that p16INK4a and p53 are positioned in distinct anticancer pathways.
p16INK4a Loss of Expression Occurs in the Setting of Promoter Methylation.
To determine the mechanism of p16INK4a inactivation in p53+/+ versus -/- tumors, MSP (11
, 18)
was performed on tumors lacking p16INK4a expression. Structural alteration of a wild-type allele of p16INK4a was seen in only 2 of 41 tumors analyzed (Table 2b
; data not shown and Ref. 11
). Of the 2 tumors in this cohort with structural alterations, 1 from a p16INK4a+/- p53+/+ mouse (#13; Fig. 2
, d and e) demonstrated loss of the wild-type p16INK4a allele by Southern analysis. The 2nd tumor, derived from a p16INK4a+/+ p53-/- animal (#21; Fig. 2e
), demonstrated an aberrantly migrating form of p16INK4a with no detectable wild-type message by reverse transcription-PCR (primers in exon 1
and 2), suggesting an intragenic mutation affecting splicing. This tumor (#21) also demonstrated promoter methylation, suggesting biallelic targeting of p16INK4a (see below).
The infrequent detection of gene deletion in tumors lacking p16INK4a expression suggested the possibility of gene silencing through promoter methylation, which has been noted previously to occur in primary murine tumors (11
, 19)
. Therefore, 17 tumors (10 p53+/+ and 7 p53-/-) were analyzed (Fig. 2e
and Table 2b
) using an MSP described previously (11
, 18)
. Of 16 tumors without detectable p16INK4a expression, 8 showed evidence of promoter methylation in bisulfate-treated DNA from primary tumors. Methylation was not detected in a tumor with deletion of the wild-type p16INK4a allele (tumor #13; Fig. 2e
) or in a tumor with detectable p16INK4a expression (tumor #15). It is worth noting that the PCR strategy will not amplify a product from the null allele (e.g., tumor #8) because the 3' primer is located in the targeted deletion of exon 1 (11)
. These data suggest that loss of p16INK4a occurs more readily through promoter methylation than deletion (Table 2b)
, even in the p53-/- setting in which the occurrence of chromosome loss and genome-wide interstitial deletions have been purported to be more frequent (17
, 20
, 21) .
The frequencies of p16INK4a loss of expression, promoter methylation, and LOH in tumors from this work and a previous analysis (11)
are compiled in Table 2b
, indicating a significant predilection for methylation over LOH (P < 0.0001), in tumors where p16INK4a expression is lost. A caveat to this analysis is that the majority of tumors included in Table 2b
were from p16INK4a+/- mice (27 of the 34 tumors in the p16INK4a loss column), and it is possible that the frequency of LOH would be higher in tumors from p16INK4a+/+ mice. Additionally, 50% of tumors with undetectable p16INK4a expression did not demonstrate either LOH or methylation, suggesting either that some other mechanism of silencing of p16INK4a occurs or that our MSP strategy is unable to detect partial promoter methylation.
|
Discussion |
|---|
|
TopABSTRACTIntroductionMaterials and MethodsResultsDiscussionREFERENCES |
|---|
Furthermore, although p16INK4a expression varied only modestly in cultures of differing densities (Fig. 1c)
, p19ARF levels varied tremendously. Previously, several groups have demonstrated p19ARF (or p14ARF) regulation by E2F (13)
, and the observation of lower levels of p19ARF in cells passaged at highest density likely reflects decreased proliferation. As cultures passaged at intermediate density grew most rapidly (Fig. 1c)
, some feature of low density besides increased proliferation/passage number appears to induce p19ARF expression. A possible interpretation of these results is that the two products of the Ink4a/Arf locus are regulated differentially by signals stemming from cell-cell interactions and/or paracrine factors. Elucidating the molecular nature of these density-mediated signals, and determining if they are the same signals that mediate p16INK4a induction and ultimately loss, in nascent tumors may lead to an improved understanding of this barrier to cancer.
Lastly, our data permit an in vivo quantification of the preference of p16INK4a loss through promoter methylation versus gene deletion. Table 2b
represents pooled data from this work and a previous analysis of carcinogen-induced tumors (11)
. These observations suggest that methylation of p16INK4a is 10-fold more probable than LOH, even in p53-/- mice, where LOH occurs with increased frequency through chromosome loss or interstitial deletion (17
, 20 , 21)
. Further support for this observation comes from the fact that p16INK4a methylation as a selection event can be observed in cultured primary murine macrophages or human mammary keratinocytes (24
, 25)
. These in vitro data suggest that p16INK4a induction, in response to culture-related stimuli, limits growth of these cell types. Our in vivo data suggest that corresponding cancer-related stimuli leads to p16INK4a induction in nascent tumors. Selection for clones with methylation-dependent silencing of p16INK4a in either case does not appear to require transformation per se but rather reflects a limit to growth imposed by p16INK4a. Our data are consistent with the high frequency of p16INK4a methylation seen in human cancers and suggests p16INK4a+/- mice provide a useful platform for the study of this process.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Supported by grants from the NIH and American Cancer Society. R. A. D. is an American Cancer Society Research Professor. 
2 To whom requests for reprints should be addressed, at Dana-Farber Cancer Institute, Harvard Medical School, 44 Binney Street (M413), Boston, MA 02115. Phone: (617) 632-6085; Fax: (617) 632-6069; E-mail: ron_depinho{at}dfci.harvard.edu
3 The abbreviations used are: MEF, murine embryo fibroblast; LOH, loss of heterozygosity; MSP, methylation-specific PCR; Rb, retinoblastoma; T-Ag, T antigen.
Received 1/15/02. Accepted 3/22/02.
|
REFERENCES |
|---|
|
TopABSTRACTIntroductionMaterials and MethodsResultsDiscussionREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. Yamakoshi, A. Takahashi, F. Hirota, R. Nakayama, N. Ishimaru, Y. Kubo, D. J. Mann, M. Ohmura, A. Hirao, H. Saya, et al. Real-time in vivo imaging of p16Ink4a reveals cross talk with p53 J. Cell Biol., August 10, 2009; 186(3): 393 - 407. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. F. Leong, J. F. L. Chau, and B. Li p53 Deficiency Leads to Compensatory Up-Regulation of p16INK4a Mol. Cancer Res., March 1, 2009; 7(3): 354 - 360. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. L. Vormer, F. Foijer, C. L. C. Wielders, and H. te Riele Anchorage-Independent Growth of Pocket Protein-Deficient Murine Fibroblasts Requires Bypass of G2 Arrest and Can Be Accomplished by Expression of TBX2 Mol. Cell. Biol., December 15, 2008; 28(24): 7263 - 7273. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. R. Ramsey, J. Krishnamurthy, X.-H. Pei, C. Torrice, W. Lin, D. R. Carrasco, K. L. Ligon, Y. Xiong, and N. E. Sharpless Expression of p16Ink4a Compensates for p18Ink4c Loss in Cyclin-Dependent Kinase 4/6-Dependent Tumors and Tissues Cancer Res., May 15, 2007; 67(10): 4732 - 4741. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Chin, L. A. Garraway, and D. E. Fisher Malignant melanoma: genetics and therapeutics in the genomic era. Genes & Dev., August 15, 2006; 20(16): 2149 - 2182. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. S. Nadella and L. S. Kirschner Disruption of Protein Kinase A Regulation Causes Immortalization and Dysregulation of D-Type Cyclins Cancer Res., November 15, 2005; 65(22): 10307 - 10315. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Ward, N. P. Nikolov, J. R. Tschetter, J. B. Kopp, F. J. Gonzalez, S. Kimura, and R. M. Siegel Progressive Glomerulonephritis and Histiocytic Sarcoma Associated with Macrophage Functional Defects in CYP1B1-Deficient Mice Toxicol Pathol, October 1, 2004; 32(6): 710 - 718. [Abstract] [PDF]
|
|
|
|
 |
|
 Z. Sachs, N. E. Sharpless, R. A. DePinho, and N. Rosenberg p16Ink4a Interferes with Abelson Virus Transformation by Enhancing Apoptosis J. Virol., April 1, 2004; 78(7): 3304 - 3311. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. Ward and D. E. Devor-Henneman Mouse Models of Human Familial Cancer Syndromes Toxicol Pathol, January 1, 2004; 32(1_suppl): 90 - 98. [Abstract] [PDF]
|
|
|
|
|
|
F. Zindy, L. M. Nilsson, L. Nguyen, C. Meunier, R. J. Smeyne, J. E. Rehg, C. Eberhart, C. J. Sherr, and M. F. Roussel Hemangiosarcomas, Medulloblastomas, and Other Tumors in Ink4c/p53-null Mice Cancer Res., September 1, 2003; 63(17): 5420 - 5427. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. D'Amico, K. Wu, D. Di Vizio, A. T. Reutens, M. Stahl, M. Fu, C. Albanese, R. G. Russell, W. J. Muller, M. White, et al. The Role of Ink4a/Arf in ErbB2 Mammary Gland Tumorigenesis Cancer Res., June 15, 2003; 63(12): 3395 - 3402. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Kannan, N. E. Sharpless, J. Xu, R. C. O'Hagan, M. Bosenberg, and L. Chin Components of the Rb pathway are critical targets of UV mutagenesis in a murine melanoma model PNAS, February 4, 2003; 100(3): 1221 - 1225. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. S. Tam, T. R. Devereux, A. C. Patel, J. F. Foley, R. R. Maronpot, and T. E. Massey Perturbations of the Ink4a/Arf gene locus in aflatoxin B1-induced mouse lung tumors Carcinogenesis, January 1, 2003; 24(1): 121 - 132. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
