BRAFE600-associated senescence-like cell cycle
arrest of human naevi
Chrysiis Michaloglou1*, Liesbeth C. W. Vredeveld1*, Maria S. Soengas3*, Christophe Denoyelle3,
Thomas Kuilman1, Chantal M. A. M. van der Horst4, Donne ´ M. Majoor2, Jerry W. Shay5, Wolter J. Mooi6
& Daniel S. Peeper1
Most normal mammalian cells have a finite lifespan1, thought to
constitute a protective mechanism against unlimited prolifer-
ation2–4. This phenomenon, called senescence, is driven by telo-
mere attrition, which triggers the induction of tumour
suppressors including p16INK4a(ref. 5). In cultured cells, senes-
such oncogene-induced senescence represents a physiological
process has long been debated. Human naevi (moles) are benign
tumours of melanocytes that frequently harbour oncogenic
mutations (predominantly V600E, where valine is substituted
for glutamic acid) in BRAF7, a protein kinase and downstream
effector of Ras. Nonetheless, naevi typically remain in a growth-
arrested state for decades and only rarely progress into malig-
nancy (melanoma)8–10. This raises the question of whether naevi
undergo BRAFV600E-induced senescence. Here we show that sus-
tained BRAFV600Eexpression in human melanocytes induces
cell cycle arrest, which is accompanied by the induction of
both p16INK4aand senescence-associated acidic b-galactosidase
(SA-b-Gal) activity, a commonly used senescence marker. Validat-
ing these results in vivo, congenital naevi are invariably positive
for SA-b-Gal, demonstrating the presence of this classical senes-
cence-associated marker in a largely growth-arrested, neoplastic
human lesion. In growth-arrested melanocytes, both in vitro and
in situ, we observed a marked mosaic induction of p16INK4a,
suggesting that factors other than p16INK4acontribute to protec-
tion against BRAFV600E-driven proliferation. Naevi do not appear
to suffer from telomere attrition, arguing in favour of an
active oncogene-driven senescence process, rather than a loss
of replicative potential. Thus, both in vitro and in vivo,
BRAFV600E-expressing melanocytes display classical hallmarks
of senescence, suggesting that oncogene-induced senescence
represents a genuine protective physiological process.
Melanocyticnaevi representanintriguing humansettingwherean
activated oncogene can co-exist with long-term arrested cells. Naevi
are very common, clonal and benign tumours of cutaneous mela-
nocytes11, and frequently harbour the V600E mutation in BRAF7
(NCBI gene bank re-named the V599E mutation based on newly
available sequence data; accession number NM_004333.2; hereafter
this mutation12,13, an initial phase of naevus growth is typically
followed by a near-complete cessation of proliferative activity,
which is maintained for many decades8,9,14. Therefore, it is conceiv-
able that growth arrest of naevi results from oncogene-induced
senescence acting as an effective cellular brake against BRAFE600-
mediated oncogenic signalling.
associated BRAFE600mutant on the proliferative capacity of freshly
isolated normal human skin melanocytes. A bicistronic lentiviral
vector co-expressing BRAFE600and enhanced green fluorescent
protein (eGFP) (to monitor infection efficiency, typically approxi-
mately 90%; Fig. 1a) was used to transduce melanocytes. Short-term
expression of BRAFE600(3–7 days) led to enhanced melanocyte
proliferation that was measured by a moderate but reproducible
increase in 5-bromodeoxyuridine (BrdU) incorporation (Fig. 1b and
data not shown). However, this effect was transient—sustained
expression of BRAFE600resulted in marked cell cycle arrest. In
most (roughly 80%) of the transduced melanocytes, this was associ-
ated with an intense activity of SA-b-Gal, a marker for senescent or
stressed cultured cells as well as for aged tissues in vivo6,15(Fig. 1c).
The p16INK4aprotein is a major tumour suppressor, often highly
expressed in senescent cells in vitro and inactivated in a variety of
human cancers, including 30–70% of melanomas16. We found that
BRAFE600-expressing melanocytes had elevated levels of p16INK4a
(Fig. 1d). Notably, the staining for p16INK4awas heterogeneous, with
25–35% of the BRAFE600-expressing melanocytes showing low or
cells were arrested (and still expressed the lentiviral cassette, as
evident from eGFP expression).
As oncogene-induced senescence in vitro has been best defined in
primary diploid fibroblasts, we analysed the cellular response to
activated BRAF in this ‘reference’ context. Consistent with previous
retroviral delivery of BRAFE600into two different strains of normal
the possibility that these effects were caused by supra-physiological
levels of BRAFE600expression, we designed a system in which its
expression levels could be manipulated with considerable precision.
Cells were transduced with a mixture of retrovirus encoding
a BRAFE600expression cassette and retrovirus encoding a short
hairpin (sh)RNA targeting both endogenous wild type and ectopic
BRAFE600(for details, see Fig. 2 legend). This strategy allowed
expression of BRAFE600to levels similar to those seen in melanoma
cells (see Supplementary Fig. 2a), or to levels close to, or indis-
tinguishable from,endogenous BRAFlevels(Fig.2a,comparelanes4
1Division of Molecular Genetics and2Division of Pathology, The Netherlands Cancer Institute, Plesmanlaan 121, 1066 CX Amsterdam, The Netherlands.3Department of
Dermatology and Comprehensive Cancer Center, University of Michigan, 1500 E Medical Center Dr. Ann Arbor, Michigan 48109, USA.4Department of Plastic, Reconstructive
and Hand Surgery, Academic Medical Centre, PO Box 22660 G4-226, 1100 AZ Amsterdam, The Netherlands.5Department of Cell Biology and Harold Simmons Cancer Center,
The University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, Texas 75390, USA.6Department of Pathology, Free University Medical Centre,
De Boelelaan 1117, 1081 HV Amsterdam, The Netherlands.
*These authors contributed equally to this work.
Vol 436|4 August 2005|doi:10.1038/nature03890
© 2005 Nature Publishing Group
The elevation of p16INK4alevels was maintained on progressively
see upregulation of p14ARFby BRAFE600(see Supplementary Fig. 3).
Furthermore, low levels of BRAFE600caused inhibition of both DNA
replication (see Supplementary Fig. 2b) and cellular proliferation
(Fig. 2b). This arrest was stably maintained, without any significant
escape (see Supplementary Fig. 2c). However, it was bypassed by co-
expression of the SV40 large tumour antigen (see Supplementary
Fig. 2d), arguing against an aphysiological effect and suggesting that
this arrest depends on cellular tumour suppressors. Furthermore, in
the context of fibroblasts defective for p53 and p16INK4aand expres-
sing SV40 small tumour antigen and human telomerase reverse
transcriptase (hTERT), BRAFE600could efficiently substitute for
RasV12in the induction of tumours in immunocompromised mice
(see Supplementary Table 1).
Recently, overexpression of RasV12has been shown to cause
accumulation of senescence-associated heterochromatic foci
(SAHF), concentrated spots of transcriptionally silenced DNA19.
We observed that low levels of BRAFE600also induced SAHF
(Fig. 2c). This was accompanied by focal accumulation of a specific
heterochromatin-associated histone modification, namely methyl-
ation of lysine 9 of histone H3 (K9M-H3). Together, these results
indicate that low levels of BRAFE600induce senescence-like cell cycle
arrest in primary human cells.
To investigate whether p16INK4aupregulation constitutes a
protective response to inappropriate mitogenic signalling by
BRAFE600, we created cell lines stably expressing p16INK4ashRNA.
This shRNA was effective, as it suppressed the accumulation of
p16INK4ain various settings (Fig. 2a), caused increased proliferation
Figure 1 | Sustained expression of BRAFE600induces a senescence-like
arrestin normalhumanmelanocytes. a,Infectionefficiency is estimatedto
be approximately 90% by visualization of eGFP-positive cells. Nuclear
BRAFE600on melanocyte proliferation measured as the percentage of
positive cells after a 3-h pulse with BrdU. DAPI stain to detect nuclei (top
panel) and BrdU-positive cells (bottom panel). Control melanocytes
remained proliferative and viable. c, SA-b-Gal activity after brief and
sustained BRAFE600expression. d, Heterogeneous levels of p16INK4a(upper
panels and lower panels, red) in BRAFE600-transduced melanocytes. DAPI
stain is used to detect nuclei (lower panels, blue). Note that at 21 days after
infection, ,1% of the BRAFE600-expressing melanocytes incorporated
Numbers given are representative of three independent experiments.
normalhumanfibroblastsina p16INK4a-independentmanner. a–c,Human
BJhTERTcells stably expressing either control or p16INK4ashRNA were
transduced with a mixture of a retrovirus encoding BRAFE600expression
shRNA targeting both endogenous wild type and ectopic BRAFE600and
blasticidin resistance (labelled BRAFE600). Pharmacological selection for
both resistance markers was used to create cells that had stably integrated
both the BRAFE600expression cassette and the shRNA. The latter alone did
not affect proliferative potential (data not shown). As a control, cells were
transduced with p16INK4a-encoding retrovirus. a, Analysis by western
blotting with b-actin used as a loading control. b, Analysis by proliferation
curves. Shown are the results of three independent experiments performed
in duplicate with standard deviations. c, DAPI staining to detect SAHF
(percentage of SAHF-positive cells is indicated in insert) and
immunofluorescence for K9M-H3. Note that whereas p16INK4ashRNA
bypass BRAFE600-induced senescence. Even undetectable levels of BRAFE600
in the absence of p16INK4a.
NATURE|Vol 436|4 August 2005
© 2005 Nature Publishing Group
(see Supplementary Fig. 4) and abolished the induction of cell cycle
arrest by overexpressed p16INK4a(Fig. 2b; see also Supplementary
Fig. 2b, c). However, low levels of BRAFE600inhibited proliferation
and induced SAHF even on p16INK4adepletion (Fig. 2b, c; see also
Supplementary Fig. 2b, c). Identical observations were made for a
p16INK4abinding-deficient cyclin-dependent kinase 4 mutant
(CDK4R24C; see Supplementary Fig. 2e), for a different fibroblast
strain (TIG3; see Supplementary Fig. 2f–h) and for fibroblasts
explanted from a homozygous p16INK4a-deficient ‘Leiden’ patient20
(see Supplementary Fig. 2i–l). These results suggest that, at least in
cultured human fibroblasts, p16INK4ais upregulated in response to
physiological levels of BRAFE600, but is not strictly required for the
induction of cell cycle arrest.
Next, we wished to validate in vivo our observation that BRAFE600
induces senescence-like arrest in vitro. Given the high frequency of
BRAF mutations in both naevi and melanomas7,12, we looked for
hallmarks of senescence in a panel of resection specimens of human
naevi. We first confirmed the presence of the BRAFE600mutation in
eight specimens of our panel of 23 naevi (see Supplementary Fig. 5).
We then used paraffin-embedded as well as cryosections of normal
human skin and resection specimens of naevi for further analysis.
Virtually all melanocytes within these naevi were growth-arrested, as
judged by negative immunohistochemical staining for the prolifer-
ation marker Ki-67 (Fig. 3a), in agreement with the literature8,9. This
was in contrast to epidermal keratinocytes in the same specimens,
many of which showed proliferative activity.
As expression of BRAFE600in cultured human melanocytes led to
cell cycle arrest along with induction of SA-b-Gal activity, we
subjected our naevi panel to analysis for this senescence marker.
SA-b-Gal activity has previously been shown to increase in human
epidermis as a function of age15. To avoid detection of age-induced
SA-b-Gal activity, we analysed a panel of congenital naevi obtained
from patients under one year of age. Indeed, all 23 naevi analysed
displayed high levels of SA-b-Gal activity (Fig. 3a–c). This activity
was absent from normal skin melanocytes within the same samples
(Fig. 3b, c) and in control samples taken from normal skin (Fig. 3a).
In agreement with previous data21, SA-b-Gal activity was also absent
Terminally differentiated keratinocytes within the epidermis of the
same samples also did not show SA-b-Gal activity (Fig. 3b, c,
arrowheads), consistent with previous observations made in young
(,39yr of age) donors15. SA-b-Gal activity was similarly lacking
from the cells of skin adnexa, such as the sweat gland (Fig. 3b, c,
labelled S), and from the dermal mesenchymal cells between the
naevus cell nests. Thus, human naevi, largely growth-arrested neo-
plastic lesions, are positive for the senescence marker SA-b-Gal.
Studies in mouse models and humans indicate that (epi)genetic
inactivation of the p16INK4agene is associated with melanomagen-
esis9,16,22,23. Normal melanocytes, scattered alongside the dermal–
epidermal junction and surrounded by keratinocytes, had undetect-
able levels of p16INK4a(Fig. 3a). In contrast, naevi invariably con-
tained p16INK4a-expressing cells, in agreement with previous
observations9,22. Ofnote,we failedto detectany significantupregula-
tion ofp53andp21CIP1innaevi(data notshown). Irrespective ofthe
mutational status of BRAF in the naevi, the percentage of p16INK4a-
positivecellsand the intensityof staining percellwas heterogeneous,
with a striking mosaic pattern of p16INK4aimmunopositivity seen in
most naevi (Fig. 3a; see also Supplementary Fig. 6; data not shown).
Figure 3 | Human melanocytic naevi display the hallmarks of senescent
cells. Paraffin-embedded sections of human naevi and normal skin were
subjectedto immunohistochemistrywiththe indicatedantibodies. a,Melan
A (brown) identifies melanocytes, MIB1 (brown) recognizes the
a–c, Frozen sections of human naevi were subjected to SA-b-Gal staining.
The blue staining corresponds exactly to the sites of naevus cell nests. All
cells of the epidermis (mostly keratinocytes, arrowheads) and sweat gland
(S), as well as dermal cells between the naevus cell nests, are completely
negative for SA-b-Gal (b and c). Note that the brown staining in these
samples comes from pigment released by the naevus cells. Where indicated,
haematoxylinandeosin (HE) wasusedasa counter stain onthe consecutive
(SA-b-Gal-stained) section to visualize the melanocyte lesions, within the
context of the surrounding tissue.
NATURE|Vol 436|4 August 2005
© 2005 Nature Publishing Group
Importantly, all melanocytes within these p16INK4a-mosaic naevi
(whether positive or negative for p16INK4a) were growth arrested
(Fig. 3a; see also Supplementary Fig. 6). Indeed, the lack of pro-
liferation of the naevus melanocytes was associated with positive
SA-b-Gal staining rather than with p16INK4aimmunoreactivity,
thereby closely mimicking BRAFE600-expressing melanocytes in vitro.
As naevi are thought to be clonal11, it is unlikely that a differential
BRAF mutation pattern underlies the p16INK4amosaicism. The
observed p16INK4amosaicism is in accordance with the occurrence
of growth arrest in naevi of homozygous p16INK4a-deficient Leiden
patients20. However, the naevi of these patients are increased in
number and size compared to those of p16INK4a-proficient indi-
expressing, growth-arrested melanocytes in vitro, these observations
suggest that p16INK4acollaborates with other, yet-to-be-identified
factors in the establishment of long-term growth arrest of naevi.
Whether p16INK4acontributes to an irreversible arrest that, once
established, no longer requires p16INK4a(as has been suggested
previously19,24), or is involved in aspects of melanoma progression
other than oncogene-induced senescence remains to be determined.
Together these results show that human naevi and BRAFE600-
expressing melanocytes share three senescence-associated markers:
stable growth arrest, heterogeneous induction of p16INK4aand SA-b-
Although these features support oncogene-induced senescence as
the cause for cell cycle arrest of melanocytic naevi, in principle these
senescence markers may have been triggered by telomere attrition,
Indeed, telomere attrition has been previously proposed as a major
factor involved in the cell-cycle arrest of naevi25. This would be
consistent with the clinical observation that congenital naevi (that is,
arising before birth) can cover large areas of the body surface, in
contrast to naevi acquired later in life, which are usually much
cent in situ hybridization) on congenital naevus resections. As
expected, melanoma metastases, stained as controls, displayed
much less fluorescence than their surrounding stromal cells, indi-
telomere length in situ (Fig. 4a, b; see also Supplementary Fig. 7). In
contrast, we did not observe any significant difference in telomere
fluorescence when comparing congenital naevi to surrounding
tissues, in agreement with previous observations made in acquired
and Spitz naevi26. Although we cannot formally rule out the possi-
response, one would not expect the other telomeres to remain
apparently full length. These results argue in favour of an active
oncogene-driven senescence process, rather than senescence trig-
gered by exhaustion of replicative potential resulting from gross
Our observation that BRAFE600initially stimulates moderate
melanocyte proliferation supports the hypothesis that it contributes
to the initiating events of melanomagenesis. However, our results
suggest that, both in vitro and in vivo, oncogenic BRAF signalling
subsequently leads to a growth-inhibitory response, which is associ-
ated with the known classical hallmarks of senescence (that is, stable
proliferative arrest, an increase in p16INK4aand the induction of SA-
b-Gal activity). Our results therefore provide support for a specu-
lative model previously proposed9,27, inwhich BRAFE600cannot fully
transform human melanocytes, but requires additional, cooperating
events for tumour development. Supporting this view, zebrafish
expressing a BRAFE600transgene develop ‘fish-naevi’, which require
observations provide evidence of oncogene-induced senescence as a
physiological mechanism in humans limiting the progression of
Plasmids. Gene transfer in normal human melanocytes was achieved with two
independent lentiviral vectors (FG12-HA-BRAFE600-eGFP and HIV-CS-CG-
BRAFE600-puro), with comparable phenotypes. pBabe-puro-BRAFE600was
used to transduce fibroblasts.
The p16INK4ashRNA sequence corresponds to nucleotides 21–41 (GenBank
accession number NM_000077). The BRAF shRNA sequence was previously
Cell culture, retroviral transduction, cell cycle analysis and proliferation
curves. Normal human melanocytes were isolated from the epidermis of
neonatal foreskins. Briefly, foreskins were incubated overnight in trypsinization
solution (22.5mM HEPES, 7.5mM glucose, 2.25mM KCl, 100mM NaCl,
0.75mM Na2HPO4and 0.17% trypsin). Dermis and epidermis were separated
by scraping. The epidermal compartment was incubated for 3–4d in 254CF
medium (Cascade Biologicals) supplemented with 0.1mM CaCl2, 2% fetal
bovine serum and keratinocyte growth supplement (Cascade Biologicals).
Melanocytes were subsequently separated from keratinocytes by differential
immunostaining for HMB-45. For long-term cultures, melanocytes were pro-
pagated in 254CF medium in the presence of 0.2mM CaCl2and melanocyte
growth supplement (Cascade Biologicals).
BJ cells expressing hTERT (in pBabeHygro) were grown in DMEM/medium
199 (Gibco) in a 4:1 ratio supplemented with 15% fetal bovine serum (PAA
Laboratories), 0.1mM MEM non-essential amino acids (Gibco), 2mM gluta-
mine (Gibco), 100unitsml21penicillin and 0.1mgml21streptomycin.
Lentiviral infections were performed using HEK293T cells as producers of
viral supernatants. The Phoenix packaging cell line was used for the generation
of ecotropic retroviruses, as described30. Amphotropicretroviruses encodingthe
Figure 4 | No apparent telomere loss in naevi. a, Frozen sections of human
naevi and melanomas were subjected to telomere FISH (red). A
representative sample is shown; six naevi were analysed (five of which
green) and DAPI (blue). Note that in the naevus specimen, both the naevus
cells (Melan-A positive) and the stromal compartment (Melan-A negative)
are FISH positive. In contrast, melanoma tumour cells, which conceivably
have undergone telomere attrition, have a much weaker telomere
fluorescence signal than their surrounding stromal cells. The channel
overlay is shown (individual channels are shown in Supplementary Fig. 7).
their respective stromal cells (non-melanocytes) with standard deviations.
NATURE|Vol 436|4 August 2005
© 2005 Nature Publishing Group
ecotropic receptor were generated in HEK293T cells. All infections were carried Download full-text
out as described30.
BrdU labelling was carried out for 3h in reduced growth factor conditions.
Proliferation curves were performed as described6,30.
Human tissue samples. Surgical resection specimens of congenital naevi were
obtained from patientsin their first year of life. Specimens were freshfrozen and
stored for a period of 2months to 2years at 2708C before use.
Analysis of SA-b-galactosidase activity. Tissues were fixed in 4% formaldehyde
Na2HPO4, pH 6.0 and stained as described15. Tissues were post-fixed overnight
in 4% formaldehyde and embedded in paraffin. Cultured cells were stained as
Immunohistochemistry and antibodies. Tissues were fixed in 4% formal-
dehyde overnight and embedded in paraffin. Paraffin sections were deparaffi-
nized, rehydrated, incubated in 0.1mM sodium citrate pH 6.0, washed and
incubated with peroxidase blocking reagent (S2001; DAKO). The tissue was
incubated with the primary antibodies Ab-3 for Melan A (Molecular
Probes), MIB-1 for Ki-67 (DAKO) and Ab-4 (16P04) for p16INK4a(Molecular
Probes). Secondary antibody used was PowerVision þ (DPVB þ 999HRP;
ImmunoLogic). Peroxidase activity was detected with Liquid DAB (K3468;
The antibodies used for western blotting were F-7 for BRAF (sc-5284; Santa
Cruz), JC8 (MS-889; NeoMarkers) for p16INK4a, C-22 for CDK4 (sc-260; Santa
Cruz) and AC-74 for b-actin (A5316; Sigma). The antibodies used for immuno-
fluorescence were anti-trimethyl-histone H3 (Lys9) (07-442; Upstate) and
554070 (BD Pharmingen) for p16INK4a. DAPI staining was used to visualize
Telomere FISH. Tissue sections were prepared for FISH analysis with the
used and also the DNA staining dye DAPI. Images of fluorescein (FITC),
TAMRA and DAPI fluorescence were acquired on a digital image microscopy
software) for all individual telomeres within 12 nuclei per specimen, both for
melanocytes (within naevi and melanomas) and for cells within the stromal
compartments of both types of lesion, for six independent naevi and three
0, specific for telomeric sequences. The Melan A antibody Ab-3 was
0phosphoramidate probe: 5
Received 19 January; accepted 8 June 2005.
1. Hayflick, L. The limited in vitro lifetime of human diploid cell strains. Exp. Cell
Res. 37, 614– -636 (1965).
Mathon, N. F. & Lloyd, A. C. Cell senescence and cancer. Nature Rev. Cancer 1,
203– -213 (2001).
Lowe, S. W., Cepero, E. & Evan, G. Intrinsic tumour suppression. Nature 432,
307– -315 (2004).
Campisi, J. Senescent cells, tumour suppression, and organismal aging: good
citizens, bad neighbors. Cell 120, 513– -522 (2005).
Shay, J. W. & Roninson, I. B. Hallmarks of senescence in carcinogenesis and
cancer therapy. Oncogene 23, 2919– -2933 (2004).
Serrano, M., Lin, A. W., McCurrach, M. E., Beach, D. & Lowe, S. W. Oncogenic
ras provokes premature cell senescence associated with accumulation of p53
and p16INK4a. Cell 88, 593– -602 (1997).
Pollock, P. M. et al. High frequency of BRAF mutations in nevi. Nature Genet. 33,
19– -20 (2003).
Kuwata, T., Kitagawa, M. & Kasuga, T. Proliferative activity of primary
cutaneous melanocytic tumours. Virchows Arch. A Pathol. Anat. Histopathol.
423, 359– -364 (1993).
Bennett, D. C. Human melanocyte senescence and melanoma susceptibility
genes. Oncogene 22, 3063– -3069 (2003).
10. Chin, L., Merlino, G. & DePinho, R. A. Malignant melanoma: modern black
plague and genetic black box. Genes Dev. 12, 3467– -3481 (1998).
11.Robinson, W. A. et al. Human acquired naevi are clonal. Melanoma Res. 8,
499– -503 (1998).
12. Davies, H. et al. Mutations of the BRAF gene in human cancer. Nature 417,
13. Wellbrock, C. et al. V599EB-RAF is an oncogene in melanocytes. Cancer Res.
64, 2338– -2342 (2004).
14. Mooi, W. J. & Krausz, T. Biopsy Pathology of Melanocytic Disorders 56– -105
(Chapman & Hall Medical, London, 1992).
15. Dimri, G. P. et al. A biomarker that identifies senescent human cells in culture
and in aging skin in vivo. Proc. Natl Acad. Sci. USA 92, 9363– -9367 (1995).
16. Sharpless, E. & Chin, L. The INK4a/ARF locus and melanoma. Oncogene 22,
3092– -3098 (2003).
17. Zhu, J., Woods, D., McMahon, M. & Bishop, J. M. Senescence of human
fibroblasts induced by oncogenic Raf. Genes Dev. 12, 2997– -3007 (1998).
18. Lin, A. W. et al. Premature senescence involving p53 and p16 is activated in
response to constitutive MEK/MAPK mitogenic signalling. Genes Dev. 12,
3008– -3019 (1998).
19. Narita, M. et al. Rb-mediated heterochromatin formation and silencing of E2F
target genes during cellular senescence. Cell 113, 703– -716 (2003).
20. Gruis, N. A. et al. Homozygotes for CDKN2 (p16) germline mutation in Dutch
familial melanoma kindreds. Nature Genet. 10, 351– -353 (1995).
21. Bandyopadhyay, D. et al. The human melanocyte: a model system to study the
complexity of cellular aging and transformation in non-fibroblastic cells. Exp.
Gerontol. 36, 1265– -1275 (2001).
22. Wang, Y. L., Uhara, H., Yamazaki, Y., Nikaido, T. & Saida, T.
Immunohistochemical detection of CDK4 and p16INK4 proteins in cutaneous
malignant melanoma. Br. J. Dermatol. 134, 269– -275 (1996).
23. Kamb, A. et al. A cell cycle regulator potentially involved in genesis of many
tumour types. Science 264, 436– -440 (1994).
24. Beausejour, C. M. et al. Reversal of human cellular senescence: roles of the p53
and p16 pathways. EMBO J. 22, 4212– -4222 (2003).
25. Bastian, B. C. Understanding the progression of melanocytic neoplasia using
genomic analysis: from fields to cancer. Oncogene 22, 3081– -3086 (2003).
26. Miracco, C. et al. Quantitative in situ evaluation of telomeres in fluorescence in
situ hybridization-processed sections of cutaneous melanocytic lesions and
correlation with telomerase activity. Br. J. Dermatol. 146, 399– -408 (2002).
27. Peeper, D. S. & Mooi, W. J. Pathogenesis of melanocytic naevi: growth arrest
linked with cellular senescence? Histopathology 41, S139– -S143 (2002).
28. Patton, E. E. et al. BRAF mutations are sufficient to promote nevi formation and
cooperate with p53 in the genesis of melanoma. Curr. Biol. 15, 249– -254
29. Hingorani, S. R., Jacobetz, M. A., Robertson, G. P., Herlyn, M. & Tuveson, D. A.
Suppression of BRAF(V599E) in human melanoma abrogates transformation.
Cancer Res. 63, 5198– -5202 (2003).
30. Peeper, D. S. et al. A functional screen identifies hDRIL1 as an oncogene that
rescues RAS-induced senescence. Nature Cell Biol. 4, 148– -153 (2002).
Supplementary Information is linked to the online version of the paper at
Acknowledgements We thank D. Atsma, E. Mesman and J. Zevenhoven for help
with immunohistochemistry; S. Douma for analytical support; L. Oomen, L.
Brocks and J. van Rheenen for help with microscopy; N. Gruis and C. Out for
p16INK4a-deficient fibroblasts; L. Zaal and A. van der Wal for help with obtaining
congenital naevus specimens; M. Voorhoeve and R. Agami for pRetroSuper,
pRetroSuper-Blasticidin and pRetroSuper-GFP; S. Gryaznov for the telomere
probe; R. Beijersbergen and M. van Lohuizen for reagents; G. Abou-Rjaily for
help with lentiviral infections; P. Krimpenfort and colleagues in the Peeper
laboratory for discussions; R. Bernards for support; and M. van Lohuizen and
A. Berns for suggestions and reading of the manuscript. M.S.S is supported by
an NIH grant. M.S.S. is a V Foundation for Cancer Research Scholar. L.C.W.V.,
T.K. and D.S.P. were supported by the Netherlands Organization for Scientific
Author Information Reprints and permissions information is available at
npg.nature.com/reprintsandpermissions. The authors declare no competing
financial interests. Correspondence and requests for materials should be
addressed to D.S.P. (email@example.com) or W.J.M. (firstname.lastname@example.org).
NATURE|Vol 436|4 August 2005
© 2005 Nature Publishing Group