SEARCH

SEARCH BY CITATION

Keywords:

  • Cdk4;
  • mouse models;
  • melanoma;
  • p53;
  • naevi;
  • Arf

Summary

  1. Top of page
  2. Summary
  3. Significance
  4. Introduction
  5. Methods
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References
  10. Supporting Information

We report on a systematic analysis of genotype-specific melanocyte (MC) UVR responses in transgenic mouse melanoma models along with tumour penetrance and comparative histopathology. pRb or p53 pathway mutations cooperated with NrasQ61K to transform MCs. We previously reported that MCs migrate from the follicular outer root sheath into the epidermis after neonatal UVR. Here, we found that Arf or p53 loss markedly diminished this response. Despite this, mice carrying these mutations developed melanoma with very early age of onset after neonatal UVR. Cdk4R24C did not affect the MC migration. Instead, independent of UVR exposure, interfollicular dermal MCs were more prevalent in Cdk4R24C mice. Subsequently, in adulthood, these mutants developed dermal MC proliferations reminiscent of superficial congenital naevi. Two types of melanoma were observed in this model. The location and growth pattern of the first was consistent with derivation from the naevi, while the second appeared to be of deep dermal origin. In animals carrying the Arf or p53 defects, no naevi were detected, with all tumours ostensibly skipping the benign precursor stage in progression.


Significance

  1. Top of page
  2. Summary
  3. Significance
  4. Introduction
  5. Methods
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References
  10. Supporting Information

Our data suggest that the p53 pathway is important in regulating the ‘stress’ response of normal melanocytes (MCs). In contrast, activated Cdk4 does not affect the MC UVR response, but instead induces the formation of dermal naevus-like lesions in mice. Thus activated Cdk4 appears to support naevus development, whereas Arf or p53 loss may result in its bypass. These data provide in vivo functional evidence in support of a role for cell cycle regulatory pathways in controlling naevogenesis and in determining whether melanomas develop from such benign precursors or arise de novo.

Introduction

  1. Top of page
  2. Summary
  3. Significance
  4. Introduction
  5. Methods
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References
  10. Supporting Information

Malignant melanoma (MM) susceptibility is characterized by mutations of the CDKN2A gene, which encodes two transcripts. The first, p16INK4A (INK4A) is a CDK-inhibitor that binds to and inhibits CDK4 (and CDK6), which function via phosphorylating pRb and inducing G1-S phase cell cycle progression. The second, alternatively spliced CDKN2A transcript, p14ARF (ARF), acts through a distinct pathway involving stabilization of p53 through abrogation of MDM2-induced p53 degradation (Zhang et al., 1998), albeit with p53-independent functions of ARF becoming increasingly evident (Sherr, 2006). Germline CDKN2A mutations most often affect INK4A, or both transcripts, but ARF-specific mutations also predispose to MM (e.g. Randerson-Moor et al., 2001; Rizos et al., 2001). ARF can also be frequently lost during MM tumour progression (Freedberg et al., 2008).

In addition to CDKN2A, there is a second MM susceptibility locus, CDK4, with some MM kindreds carrying germline mutations such as R24C and R24H (e.g. Zuo et al., 1996). Again, this mutation affects the function of the INK4A/pRb pathway, with the mutations rendering the kinase resistant to INK4A inhibition. To date, CDKN2A and CDK4 are the only confirmed familial MM susceptibility genes.

RAS/RAF/MAPK pathway activation is also critical for MM development. The MAPK pathway is activated through mutation of BRAF or NRAS in about 70% of all MMs (Hocker and Tsao, 2007). Importantly, BRAF or NRAS mutation is present in most naevi (Pollock et al., 2003), which are sometimes MM precursors (Grichnik, 2008). In terms of cutaneous lesions, BRAF mutation is not commonly seen in lesions emanating from anatomical areas of chronic UVR exposure but is often found in MMs from areas of intermittent sun exposure such as the back and trunk (Maldonado et al., 2003). Mice carrying MC-specific activating mutations in Hras (Broome Powell et al., 1999) or Nras (Ackermann et al., 2005) or Braf (Dankort et al., 2009; Dhomen et al., 2009; Goel et al., 2009) can develop melanocytic lesions spontaneously.

Mouse models harbouring mutations in genes involved in human MM can offer important insights into the function of these genes in tumorigenesis. Ink4a, Arf and p53 knockout mice rarely, if ever, develop MM, with or without neonatal UVR (Walker and Hayward, 2002). Dual Ink4a/Arf-null animals develop MM after neonatal UVR (Yang et al., 2007), although this result may be somewhat strain dependent (Serrano et al., 1996). Invariably, when these mice are crossed with transgenics carrying MC-specific activating Braf, Nras or Hras mutations, MM develops rapidly (e.g. Kannan et al., 2003; Goel et al., 2009; Ackermann et al., 2005; Hacker et al., 2006). The role of the products of the Cdkn2a locus, Ink4a and Arf, has been particularly interesting. Melanocyte-specific activation of Hras on both Ink4a- and Arf-null backgrounds leads to spontaneous MM development (Kannan et al., 2003). However, neonatal UVR increases penetrance in Arf−/−/Tyr-Hras but not Ink4a−/−/Tyr-Hras animals. Spontaneous MMs in Arf−/−/Tyr-Hras animals have a high degree of chromosomal instability, while UVR-induced lesions are cytogenetically intact except for frequent genomic amplification targeting Cdk6 (O’Hagan et al., 2003). Deletion of p53 on the same Tyr-Hras transgenic background also results in spontaneous MM development (Bardeesy et al., 2001), although the effects of neonatal UVR were not assessed. Recently, Ha et al. (2007), using the Mt-Hgf model, showed that Ink4a or Arf abrogation significantly increased MM penetrance after neonatal UVR exposure. They did not report spontaneous MM rates and had difficulty assessing the role of p53 in MM as p53−/−/Hgf mice rapidly develop rhabdomyosarcoma (Ha et al., 2007). BRAFV600E can also cooperate with p53 loss in MM induction (Goel et al., 2009). TP53 mutations are seen in up to 13% of primary human MMs, their role in MM genesis is significant, although often under recognized. A review of all known sequence variants in MM (Hocker and Tsao, 2007) showed that, in terms of point mutations, TP53 is the most frequently mutated tumour suppressor in primary MM (CDKN2A is abrogated at higher rates, predominantly because of large deletions).

While MMs develop on animals carrying MC-specific HrasG12V and NrasQ61K mutations (Broome Powell et al., 1999; Ackermann et al., 2005), age-of-onset is late, penetrance is low, and tumours are small lesions with features reminiscent of dermal naevi. Other mutations are necessary to induce overt MC malignancy in mice. In our previous study, we showed that after UVR exposure, Tyr-HrasG12V mice developed small in loco lesions reminiscent of dermal naevi, while the co-introduction of the Cdk4R24C mutation resulted in the development of very aggressive malignant lesions (Hacker et al., 2006). Here, we sought to compare the roles of the pRb and p53 axes in MM development. We found differences in the phenotype of MCs in the skin, the response of MCs to neonatal UVR and the comparative pathology of the resultant MMs.

Methods

  1. Top of page
  2. Summary
  3. Significance
  4. Introduction
  5. Methods
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References
  10. Supporting Information

Mouse melanoma models

Mouse models and genotyping have been previously described: Cdk4R24C (Rane et al., 1999), Tyr-NrasQ61K (Ackermann et al., 2005) and Tyr-Cre (ER)T2 (Bosenberg et al., 2006). Arf−/− and p53F/F mice (carrying floxed alleles allowing Cre-mediated excision of exons 1-10) were obtained from the Mouse Models for Cancer Consortium (http://mouse.ncifcrf.gov). Mice were bred for at least five generations onto an FVB background. Melanocyte-specific p53 deletion in p53F/F/Tyr-Cre(ER)/Tyr-Nras mice was induced via topical application of 8-OH-tamoxifen (15 mg/ml in DMSO) at P0, 1 and 2 (Bosenberg et al., 2006). All experiments were undertaken with institute animal ethics approval. Mice were killed before tumours exceeded 10 mm in diameter, or when they ulcerated.

UVR treatments

Pups (3-day-old) were given a 20- min exposure to UVB from a bank of six cellulose acetate-filtered Phillips (Eindhoven, The Netherlands) TL100W 12RS UVB lamps (total UVB dose 5.9 kJ/m2, or an erythemally weighted dose of 1.8 kJ/m2). UVB dose was measured using a Solar Light (Glenside, PA, USA) PMA2100 radiometer with either a PMA2101 detector to measure biologically weighted UVB or a PMA2106 detector to measure non-weighted UVB.

Immunohistochemistry

Paraffin-embedded sections of mouse skin and tumours were dewaxed and treated with Dako low-pH antigen retrieval solution (Dako, Ely, UK) at 125°C for 5 min. Endogenous peroxidase activity was quenched in 3% H202 for 10 min, and sections were washed and blocked with 10% goat serum. The primary antibodies were a p53 polyclonal antibody (CM5; Novocastra, Leica Biosystems, Newcastle Upon Tyne, UK) applied at 1:500, Pax3 polyclonal antibody (a gift from Dr Rick Sturm), diluted 1:500, and a tyrosinase-related protein 1 (Tyrp1, αPEP1) polyclonal antibody diluted 1:500. Secondary detection was via an Envision plus detection kit (Dako), which was visualized using AEC plus (Dako), and sections were counterstained with haematoxylin.

Dual label immunofluorescence

Sections were dewaxed, and antigen retrieval performed as mentioned earlier. Sections were blocked with 1% BSA and incubated with rat monoclonal anti-Ki-67 (Dako). After washing, biotinylated donkey anti-rat (Jackson ImmunoResearch Laboratories, West Grove, PA, USA) or biotinylated donkey anti-mouse (Jackson) antibody was applied (at 1:300 dilution) for 1 h, followed by washing and incubation with AlexaFluor 488-labelled streptavidin (1:300) for 1 h. For double label studies, the second primary antibody (anti-Trp1) was applied (1:300) for 1 h. AlexFluor 555-labelled donkey anti-rabbit (Jackson) was added (1:300) for 1 h. Slides were mounted with Vector Shield (Vector laboratories, Burlingame, CA, USA) containing DAPI. Slides were viewed on a fluorescent microscope, and positive cells counted. The number of basal MCs (AlexaFluor 555-labelled dendritic cells) per 40× field was counted, along the length of the skin (normally from 15 and 25 fields per skin).

Statistical analysis

The survival of mice in each treatment group was estimated using Kaplan–Meier analysis, and the log-rank test was used to test for pairwise differences between the groups. The associations between treatment and the proportion of mice that developed disease characteristics were tested using pairwise Fisher’s exact tests or Mann–Whitney U tests.

Results

  1. Top of page
  2. Summary
  3. Significance
  4. Introduction
  5. Methods
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References
  10. Supporting Information

Longitudinal study of spontaneous and UVR-induced melanomas

We assessed MM penetrance in Cdk4R24C/R24C/Tyr-Nras, Arf−/−/Tyr-Nras, and p53F/F/Tyr-Cre(ER)/Tyr-Nras mice after a single neonatal UVB dose of 5.7 kJ/m2, compared with untreated mice of the same genotype. Figure 1 shows the Kaplan–Meier curves for MM-free survival. During the period of study (300 days), we did not observe any MMs after UVR exposure in littermates of any genotype that did not carry NrasQ61K, nor in animals carrying only this mutation. NrasQ61K cooperated with Cdk4R24C, Arf−/− and MC-specific p53 deletion, to induce MM spontaneously. Penetrance was complete in all but the Arf−/−/Tyr-Nras mice, which tended to die from other causes, especially lymphoma and sarcomas (Kamijo et al., 1997) before they succumbed to MM. Notwithstanding our inability to generate spontaneous lesions in Arf−/−/Tyr-Nras mice, spontaneous MM penetrance between genotypes was not significantly different (Figure S1). However, neonatal UVR significantly decreased age of onset in all genotypes (Figure 1), particularly effectively in association with Arf nullizigosity. We studied a small number of animals (n = 6) constitutionally null for p53 (p53−/−/Tyr-Nras). These animals also rapidly developed MMs spontaneously (average age-of-onset 93 days). We did not assess the role of neonatal UVR in this setting. Because of the early onset of lymphoma and sarcoma (Jacks et al., 1994), and the relative difficulty in generating the compound genotype, we instead chose to utilize the MC-specific p53 mutant that is unaffected by non-MM morbidity.

image

Figure 1.  Melanoma penetrance. Kaplan–Meier curve showing the time to spontaneous and UVR-induced malignant melanoma (MM) development in the various genetically modified mice. Animals that died without developing MM are represented by a vertical black dash. Neonatal UVR treatment significantly increased melanoma penetrance in all strains. (A) Cdk4R24C/R24C/Nras, P < 0.0001, log rank test. (B) Arf−/−/Nras, P = 0.0001. (C) p53F/F/Tyr-Cre/Nras, P = 0.0013. (D) Comparison of UVR-induced melanoma-free survival between genotypes. Cdk4 versus Arf, P = 0.0004; Cdk4 versus Trp53, P = 0.0004, Arf versus Trp53, P = 0.9343, N.S. The age of mice was defined by the appearance of the first cutaneous MM.

Download figure to PowerPoint

Comparative pathology of melanomas

Malignant melanomas in all genotypes had the appearance of ‘nodular’ lesions, similar to ‘animal type’ MMs (although unpigmented because of the albino FVB background). We confirmed melanocytic origin of the lesions by staining for Trp1 and Pax3 (Medic and Ziman, 2010). Although normal MCs in the skin of all genotypes stained for Trp1, its expression was lost in some tumour cells. The proportion of tumour cells positive for Trp1 was greatest in animals carrying Cdk4R24C and lowest in those carrying deleted p53 (Figures 2 and 3A). There were significant differences in histopathology between genotypes. Nuclei in p53F/F/Tyr-Cre(ER)/Tyr-Nras tumours were relatively small, with a regular ovoid shape, and high levels of Ki-67 staining (Figure 3B). Cdk4R24C/R24C/Tyr-Nras and Arf−/−/Tyr-Nras tumours exhibited more nuclear pleomorphism. The cytomorphology of the latter MMs was particularly distinct, with larger and more pleomorphic nuclei, abundant cytoplasm and numerous multinucleated giant cells. The Arf−/−/Tyr-Nras tumour cells showed very low levels of Ki-67 staining (Figure 3B).

image

Figure 2.  Histopathology of melanoma sections. (A) Images of representative tumours from each of the three genotypes stained with anti-Trp1 antibody (red-brown), and with H&E. A dotted arrow shows virtually a continuous linear grouping of naevus-like melanocytes beneath the epidermis in the Cdk4R24C/R24C/Tyr-Nras mice that is not present in the other two genotypes. The deep dermal tumours are indicated by a solid arrow. Arf−/−/Tyr-Nras and p53F/F/Tyr-Cre(ER)/Tyr-Nras rarely stained for Trp1. (B) High power images of representative H&E stained tumours of each genotype denoted by the headings above panel A. Note the differences in histopathology. Nuclei in Arf−/−/Tyr-Nras tumours are very pleomorphic, with many giant tumour cells. Nuclei in p53F/F/Tyr-Cre(ER)/Tyr-Nras malignant melanomas are ovoid and regular in shape. Scale bars = 100 μM.

Download figure to PowerPoint

image

Figure 3.  Measures of Trp1 and Ki-67 staining in melanomas. For each comparison, at least 10 tumours from each genotype, and at least 10 fields per tumour, were assessed. (A). Proportion of tumour cells (per 100) positive for Trp1. (B) Proportion of tumour cells staining for Ki-67 as a measure of proliferation in melanomas from the respective genotypes.

Download figure to PowerPoint

Naevus-like precursors associated with Cdk4 activation

On histopathological examination of the tumours, the most striking difference between genotypes was the presence or absence of naevus-like proliferations (Figure 2A). Above Cdk4R24C/R24C/Tyr-Nras lesions, we nearly always observed such ‘naevi’ immediately beneath the epidermis in a band-like distribution (Figure S2). This phenotype was observed in both UVR-treated and control mice. In terms of classical morphology, these cells are reminiscent of naevus cells, particularly a superficial congenital naevus-like pattern. They showed very low levels of proliferation, with 1.3% of these cells stained for Ki-67, significantly less than in the corresponding deep dermal MMs, where 4.9% stained (P < 0.001 Mann–Whitney U-Test). Naevi were not observed above the dermal MMs in the Arf−/−/Tyr-Nras and p53F/F/Tyr-Cre(ER)/Tyr-Nras mice (n = 28 and 23, respectively), although small nests of MCs in a similar subepidermal location were occasionally observed in their normal skin (Figure S2).

Closer assessment of tumours from Cdk4R24C/R24C/Tyr-Nras mice (n = 85), including lesions generated in previous studies, revealed two distinct MM subtypes. Fifty-five per cent (47/85) were in the deep dermis, separated from the naevus cells above by layers of collagen (Figure 2). In contrast, 42% (36/85) of lesions encompassed the subepidermal location of the naevi and appeared to be ‘expanded’ naevi, although the cells now had morphological features of malignancy. This is consistent with derivation of this subtype of MMs directly from the naevus cells (Figure 4). Only 2/85 very early tumours with disorganized architecture could not be definitively assigned to either group. Serial sectioning in most cases did not resolve the issue. We cannot rule out that all MMs emanate from the naevi and that many ‘encapsulate’, giving the appearance of deep dermal derivation. The proportion of the respective tumour ‘subtypes’ was the same in spontaneous and UVR-treated cohorts (data not shown). In contrast, examination of early lesions from the other two genotypes (Arf and p53-null) failed to show any evidence of tumour derivation from the upper dermis (Figures 2 and 5). The location of all of these MMs was consistent with derivation from follicular or other deep dermal precursors. In about 10% of hair bulbs in NrasQ61K mice, MCs are ectopically located in the bulb outer root sheath (Figure 6C) and thus are already ‘deregulated’ from the controlling effects of the follicular microenvironment. There was no significant difference in age-of-onset or penetrance of spontaneous or UVR-induced MM development (by Kaplan–Meier analysis) between the deep dermal MMs and those appearing to develop from naevi (Figure S3). In terms of anatomical location of tumours, the overwhelming majority were on the back in all genotypes, with or without UVR, and whether or not the lesions were associated with naevi (Table S1).

image

Figure 4.  Putative model for cell-of-origin of Cdk4R24C/R24C/Tyr-Nras MMs. Sections were stained with Trp1 (red-brown). Yellow arrow denotes a deep dermal MM. Dotted yellow arrows denote MMs emanating from naevi. Upper panel shows a selected series of lesions that appear to have derived from naevus-like cells in a band immediately beneath the epidermal basement membrane. The lower panel shows a deep dermal lesion, separated from the upper ‘naevus’ cells by layers of collagen, suggesting that it has not emanated from the precursor lesions above. Scale bars = 100 μM. MM, malignant melanoma.

Download figure to PowerPoint

image

Figure 5.  Putative model for cell-of-origin of Trp53F/F/Tyr-Cre(ER)/Tyr-Nras. As MMs in this genotype are almost invariably negative for Trp1, we have stained here for Pax3 (red, nuclear expression). Yellow arrow denotes melanocytic lesions. ‘Early’ MMs in these mice appear to have developed either from melanocytes ectopically located in the outer root sheath of the hair bulb or from another deep dermal interfollicular precursor(s). Scale bars = 100 μM. MM, malignant melanoma.

Download figure to PowerPoint

image

Figure 6.  Comparison of UVR-induced basal melanocyte (MC) number between strains. (A) Graph shows mean epidermal basal MC count at 3 days after neonatal UVR per 40× field over the length of each section (over ∼20–25 fields per section) ±SEM. For each strain, at least three different mice from at least two different litters were evaluated. The Mann–Whitney U-test was used to assess the significance of the differences between the values for wild type mice. Black bars carry the Tyr-NrasQ61K transgene, white bars not. (B) Skin sections from mice exposed to UVR 3 days before were stained for proliferating MCs (Trp1, red) and nuclei (DAPI, blue). Filled arrows point to epidermal basal layer MCs and dotted arrow to dermal MCs. Yellow arrow - MCs in upper hair follicle. Dotted lines denote the epidermal basal layer, and filled lines outline hair follicles. (C) Images of hair bulbs in the respective genotypes showing atypical location of MCs in the outer root sheath of the bulb. This phenomenon is seen in all genotypes carrying NrasQ61K and occurs in about 10–20% of all hair bulbs. Scale bars (yellow) = 50 μM.

Download figure to PowerPoint

MC proliferation and migration after neonatal UVR

We previously showed that in wild type and Cdk4R24C/R24C/Tyr-Nras (e.g. Figure 6A) mice after neonatal UVR, MCs emanate from the upper regions of hair follicles and migrate to the epidermal basal layer, whereas in the dermal MC population, there is no increase in MC number after UVR (Walker et al., 2009; Hacker et al., 2010). Epidermal MC count at 3–5 days after neonatal UVR is a good proxy measure for this MC ‘activation’. This response was significantly dampened in Arf−/−/Tyr-Nras and p53F/F/Tyr-Cre(ER)/Tyr-Nras mice (particularly in the latter), compared to wild type and Tyr-Nras (Figure 6). Notably, Cdk4R24C did not affect this response, either alone or in combination with Tyr-Nras (Figure 6). Thus, the Arf/p53 pathway regulates the expansion and migration of upper follicular neonatal MCs after UVB exposure. We then examined the proportion of MCs expressing Ki-67 at 3–5 days after UVR, but observed no significant difference between genotypes for either the epidermal or dermal population (Figure S4A). To better understand the discrepancy between proliferation and migration, we counted MC number at P3, the time of UVR exposure. However, we saw no significant difference in epidermal MC number at P3 between genotypes (Figure S4B). These results suggest that the increased epidermal MC density after UVR is best explained by migration from the hair follicle.

Cdk4R24C/R24C/Tyr-Nras neonatal mice had a significantly higher number of dermal MCs than the other genotypes (Figure S4C). This effect is driven by cooperation between Cdk4R24C and NrasQ61K and is not related to the UVR-induced MC response. Cdk4 has a specific effect on the dermal MC population, and these are likely to be the cells that go on to form the naevi in adults.

Discussion

  1. Top of page
  2. Summary
  3. Significance
  4. Introduction
  5. Methods
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References
  10. Supporting Information

The differential roles of the Cdk4/Ink4a/pRb and Arf/p53 pathways in MM genesis have been of particular interest, as germline mutations resulting in their deregulation are found in familial MM. Mice null for Ink4a and carrying HrasG12V do not show an increase in MM penetrance after neonatal UVR, whereas those null for Arf do (Kannan et al., 2003). One should note that in nearly all mouse models, MMs develop spontaneously, and UVR could be argued to be only a tumour promoter. Certainly in our study, Cdk4R24C/R24C/Tyr-Nras and p53F/F/Tyr-Cre(ER)/Tyr-Nras mice develop MM with complete penetrance spontaneously (Figure 1), and age of onset is decreased after neonatal UVR. As Arf−/−/Tyr-Nras mice generally succumb to lymphomas and sarcomas before they develop MMs, it is difficult to compare spontaneous MM penetrance between genotypes (Figure S1). Nonetheless, in agreement with previous studies (Kannan et al., 2003; Ha et al., 2007), we confirm the profound effect of loss of Arf in exacerbating MM development after UVR. It is unclear why this is so. In vitro studies with Arf-null murine embryonic fibroblasts (Sarkar-Agrawal et al., 2004) suggest that lack of Arf results in diminution of DNA repair capacity. In addition, Arf is involved in upregulation of Xpc, a component of the nucleotide excision repair pathway necessary for removal of pyrimidine dimers (Dominguez-Brauer et al., 2009). It remains to be seen how p53 and Arf loss affect DNA repair in MCs. However, it must be kept in mind that both p53 and Arf are tumour suppressors and do not necessarily manifest their major antitumourigenic effects in the DNA damage repair period. In radiation-induced murine lymphoma genesis, the role of Arf (and p53) in tumour suppression is only manifested after the DNA damage response has subsided, and possibly only in the few cells with destabilizing mutations that are on a trajectory to malignancy (Christophorou et al., 2006).

Lack of Arf/p53 does affect some MC behaviour in the few days following UVR exposure, i.e. it suppresses the activation and migration of neonatal MCs. The UVR-responsive MCs are most likely than those in the upper hair follicle and/or in the epidermis (i.e. those in contact with keratinocytes). Paradoxically, we do not see differences between genotypes in pre-existing epidermal MC number, nor in the proportion of epidermal MCs proliferating 3 days after UVR. Nevertheless, there are more proliferating MCs in the epidermis after neonatal UVR in wt, Tyr-Nras and Cdk4R24C/R24C/Tyr-Nras mice than in the animals with defective p53, simply because there are more MCs. Another possibility we cannot rule out is a difference in MC survival that could also help explain the difference in epidermal MC number between genotypes after UVR. The ‘expansion’ of MCs appears to originate in the upper hair follicle, which contains the bulge, the location of MC stem cells. As we previously suggested (Walker et al., 2009), there are three possibilities for the increase in neonatal epidermal MCs after UVR: (i) the MCs migrating downwards into the follicle at the time of UVR simply reverse their trajectory; (ii) MC stem cells are activated to produce more ‘transit amplifying’ (TA) cells; or (iii) TA cells (or other upper follicular MCs) divide and/or migrate upwards. This study is more consistent with the second or third explanation. Which upper follicular MCs are activated is an area of ongoing investigation. Although it is reasonable to assume that this MC activation enhances MM ‘initiation’, this appears not to be the case, as suppression of this response by removal of p53 or Arf does not reduce melanoma penetrance.

p53 is well known to play a role in sensing and transducing paracrine and autocrine signals as part of the MC pigmentation response (Khlgatian et al., 2002; Cui et al., 2007; Murase et al., 2008; Box and Terzian, 2008), and the neonatal MC activation may be coupled to such responses. We assume that the role of Arf here is at least partially because of its function in regulating p53 (Sherr, 2006). As with MC-specific p53 loss, its effects in tumour suppression are likely to be mainly cell-autonomous, as several studies have shown that MC-specific deletion of p53 (this study), Ink4a (Huijbers et al., 2006) or Pten (Dankort et al., 2009), in cooperation with MC-specific oncogenes, effectively generates MM in mice. It is not necessary to have the tumour suppressor deletion in the MC microenvironment. On the other hand, the MC UVR response only occurs in the upper follicular outer root sheath and epidermis, suggesting the necessity for keratinocyte–MC inter-action. In this study, p53 loss differentially affects MCs and MM cells: in the former, it suppresses their response to UVR, whereas in the latter, it leads to increased proliferation.

Cdk4 activation appears to have a specific effect on neonatal dermal MCs and drives the formation of dermal naevi in adult mice. These dermal cells are not controlled by keratinocytes and may exhibit different responses to cytokines (Aoki et al., 2009). Studies on familial MM do not provide much evidence that CDK4 (or INK4A) are important in naevus development. While individuals in MM-prone families carrying such mutations commonly have high naevus counts and dysplastic naevi, the mutations do not always segregate with the naevus-prone phenotype (Puig et al., 1997; de Snoo et al., 2008; Goldstein et al., 2002; Pjanova et al., 2009; Molven et al., 2005). Other evidence is more positive. Carcinogen and/or UVR treatment of Cdk4R24C or Ink4a/Arf-null mice can induce naevi (Sotillo et al., 2001; van Schanke et al., 2006). In addition, CDK4 has been shown to be an independent oncogene, amplified in MMs that do not carry BRAF or NRAS mutations (Curtin et al., 2005). Genome-wide association studies of naevus susceptibility point to variants within PLA26G (a gene involved in phospholipid metabolism) on chromosome 22 and MTAP (involved in nucleoside metabolism), adjacent to CDKN2A on chromosome 9p21 (Falchi et al., 2009). It is not known how these two genes may functionally influence naevogenesis, but there is conjecture that single nucleotide polymorphisms (SNPs) in the MTAP gene may confer long-range regulation of the CDKN2A locus encoding INK4A. This would be analogous to the OCA2 gene, whose influence on eye colour is not because of OCA2 coding variants, but to remote regulation by a SNP in the adjacent gene (Sturm et al., 2008). Thus, it is not unreasonable to speculate that variation in the regulation of INK4A (or CDK4) may confer increased risk for the development of naevi. Our murine work provides in vivo evidence that this could be the case.

In this study, one subset of our MMs appears to derive de novo, and another from naevi, suggestive of distinct mechanisms of tumorigenesis. Proof of the cell of origin awaits clonal analysis experiments, such as those that have been performed for basal cell carcinoma (Youssef et al., 2010). However, support for this idea comes from a study with Mt-Ret mice (Kumasaka et al., 2010). Here, MMs develop from pre-existing benign dermal lesions, but when crossed with endothelin receptor B (Ednrb) heterozygotes, the compound mutant mice (Mt-Ret/Ednrb+/−) develop only de novo MMs. The concept of different forms of MM deriving from different MC populations is interesting given the current debate about MM derivation from different ‘stem cells’ (Grichnik, 2008; Zalaudek et al., 2008), or at least different MC populations – either in the epidermis for superficial spreading MM, the dermis for nodular MM, or the epidermis/hair follicle for lentigo MM.

Trp1 staining was lost in some tumours. One possibility could be a transdifferentiation effect (e.g. Huijbers et al., 2006), but we did not see any evidence of this histologically, with lesions in all genotypes staining for Pax3, a human MC and MM marker (Medic and Ziman, 2010). Even the murine lesions appearing to emanate from naevi had a lower proportion of Trp1-positive cells than the naevi themselves (which we assume had 100% of cells staining). Thus, lesions can lose pigmentation markers as they progress, a phenomenon that can occur in human MM (Henning et al., 2007). This can be because of promiscuous intracellular signalling in tumour cells that leads to reduced expression of tyrosinase (and presumably other pigmentation markers), because of destruction of misfolded protein in an altered metabolic environment (reviewed in Halaban, 2002).

In summary (Table 1), we have compared the role of Cdk4, Arf and p53 mutation in generating murine MM. All animals were on a uniform strain background and carrying the same oncogenic stimulus (NrasQ61K). We found that Cdk4R24C, but not loss of Arf or p53, cooperates with NrasQ61K to induce dermal naevus-like lesions. The level of MC activation in response to neonatal UVR, which affects follicular outer root sheath and epidermal (but not dermal) MCs, was significantly reduced in animals with Arf- or MC-specific p53 deletion, suggesting that a functioning p53 pathway is important for this MC response. However, the reduction in MC activation did not lower MM penetrance. Although the presence of the Cdk4R24C mutation did not affect the MC UVR response, it did increase dermal MC density in neonates, which correlated with the presence of dermal naevi in adults. We speculate that p53 (or Arf) loss enables ‘initiated’ MCs to skip the ‘naevus phase’ and to progress rapidly to overt malignancy. How lesions emanate – either from naevi or apparently de novo – is an important clinical question in dermatology, and our murine models may help shed light on these processes.

Table 1.   Summary of melanocyte (MC) UVR responses and tumour phenotype for the respective genotypes
 Cdk4/NrasArf/Nrasp53/Nras
Neonatal skin (P3) – dermal MC number********
Neonatal skin – dermal MCs 3–5 days after UVR********
Adult skin – dermal naevi******  
Neonatal skin (P3) – epidermal MC number***
Neonatal skin – epidermal/upper follicular MCs 3–5 days after UVR**********
Tumour phenotype – nuclear pleomorphism**********
Tumour phenotype – Ki-67 levels*********
Tumour phenotype – Trp1 levels******** 

Acknowledgements

  1. Top of page
  2. Summary
  3. Significance
  4. Introduction
  5. Methods
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References
  10. Supporting Information

We are grateful to Dr Rick Sturm for the Pax 3 antibody, Vince Hearing for PEP1 antibody. We thank Drs Mariano Barbacid and Marcos Malumbres for the Cdk4R24C mice. This work was funded by the NH&MRC of Australia and the Cancer Council of Queensland. GJW is the recipient of a Senior Research Fellowship from the Cancer Council of Queensland.

References

  1. Top of page
  2. Summary
  3. Significance
  4. Introduction
  5. Methods
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References
  10. Supporting Information
  • Ackermann, J., Frutschi, M., Kaloulis, K., McKee, T., Trumpp, A., and Beermann, F. (2005). Metastasizing melanoma formation caused by expression of activated NRasQ61K on an INK4a-deficient background. Cancer Res. 65, 40054011.
  • Aoki, H., Yamada, Y., Hara, H., and Kunisada, T. (2009). Two distinct types of mouse melanocyte: differential signaling requirement for the maintenance of non- cutaneous and dermal versus epidermal melanocytes. Development 136, 25112521.
  • Bardeesy, N., Bardeesy, N., Bastian, B. et al. (2001). Dual inactivation of RB and p53 pathways in RAS-induced melanomas. Mol. Cell. Biol. 21, 21442153.
  • Bosenberg, M., Muthusamy, V., Curley, D.P., Wang, Z., Hobbs, C., Nelson, B., Nogueira, C., Horner, J.W., Depinho, R., and Chin, L. (2006). Characterization of melanocyte-specific inducible Cre recombinase transgenic mice. Genesis 44, 262267.
  • Box, N.F., and Terzian, T. (2008). The role of p53 in pigmentation, tanning and melanoma. Pigment Cell Melanoma Res. 21, 525533.
  • Broome Powell, M., Gause, P.R., Hyman, P., Gregus, J., Lluria-Prevatt, M., Nagle, R., and Bowden, G.T. (1999). Induction of melanoma in TPras transgenic mice. Carcinogenesis 20, 17471753.
  • Christophorou, M.A., Ringshausen, I., Finch, A.J., Swigart, L.B., and Evan, G.I. (2006). The pathological response to DNA damage does not contribute to p53-mediated tumour suppression. Nature 443, 214217.
  • Cui, R., Widlund, H.R., Feige, E. et al. (2007). Central role of p53 in the suntan response and pathologic hyperpigmentation. Cell 128, 853864.
  • Curtin, J.A., Fridlyand, J., Kageshita, T. et al. (2005). Distinct sets of genetic alterations in melanoma. N. Engl. J. Med. 353, 21352147.
  • Dankort, D., Curley, D.P., Cartlidge, R.A., Nelson, B., Karnezis, A.N., Damsky, W.E. Jr, You, M.J., DePinho, R.A., McMahon, M., and Bosenberg, M. (2009). Braf(V600E) cooperates with Pten loss to induce metastatic melanoma. Nat. Genet. 41, 544552.
  • Dhomen, N., Reis-Filho, J.S., da Rocha Dias, S., Hayward, R., Savage, K., Delmas, V., Larue, L., Pritchard, C., and Marais, R. (2009). Oncogenic Braf induces melanocyte senescence and melanoma in mice. Cancer Cell 15, 294303.
  • Dominguez-Brauer, C., Chen, Y.J., Brauer, P.M., Pimkina, J., and Raychaudhuri, P. (2009). ARF stimulates XPC to trigger nucleotide excision repair by regulating the repressor complex of E2F4. EMBO Rep. 10, 10361042.
  • Falchi, M., Bataille, V., Hayward, N.K. et al. (2009). Genome-wide association study identifies variants at 9p21 and 22q13 associated with development of cutaneous nevi. Nat. Genet. 41, 915919.
  • Freedberg, D.E., Rigas, S.H., Russak, J. et al. (2008). Frequent p16-independent inactivation of p14ARF in human melanoma. J. Natl Cancer Inst. 100, 784795.
  • Goel, V.K., Ibrahim, N., Jiang, G., Singhal, M., Fee, S., Flotte, T., Westmoreland, S., Haluska, F.S., Hinds, P.W., and Haluska, F.G. (2009). Melanocytic nevus-like hyperplasia and melanoma in transgenic BRAFV600E mice. Oncogene 28, 22892298.
  • Goldstein, A.M., Chidambaram, A., Halpern, A., Holly, E.A., Guerry, D., Sagebiel, R., Elder, D.E., and Tucker, M.A. (2002). Rarity of CDK4 germline mutations in familial melanoma. Melanoma Res. 12, 5155.
  • Grichnik, J.M. (2008). Melanoma, nevogenesis, and stem cell biology. J. Invest. Dermatol. 128, 23652380.
  • Ha, L., Ichikawa, T., Anver, M. et al. (2007). ARF functions as a melanoma tumor suppressor by inducing p53-independent senescence. Proc. Natl Acad. Sci. USA 104, 1096810973.
  • Hacker, E., Muller, H.K., Gabrielli, B., Lincon, D., Pavey, S., Hayward, N., and Walker, G. (2006). Spontaneous and UVR-induced multiple metastatic melanoma in Cdk4R24C/R24C/TPras mice. Cancer Res. 66, 29462952.
  • Hacker, E., Muller, H.K., Hayward, N., Fahey, P., and Walker, G. (2010). Enhancement of DNA repair using topical T4 endonuclease V does not inhibit melanoma formation in Cdk4R24C/R24C/Tyr-NrasQ61K mice following neonatal UVR. Pigment Cell Melanoma Res. 23, 121128.
  • Halaban, R. (2002). Pigmentation in melanomas: changes manifesting underlying oncogenic and metabolic activities. Oncol. Res. 13, 38.
  • Henning, J.S., Dusza, S.W., Wang, S.Q., Marghoob, A.A., Rabinovitz, H.S., Polsky, D., and Kopf, A.W. (2007). The CASH (color, architecture, symmetry, and homogeneity) algorithm for dermoscopy. J. Am. Acad. Dermatol. 56, 4552.
  • Hocker, T., and Tsao, H. (2007). Ultraviolet radiation and melanoma: a systematic review and analysis of reported sequence variants. Hum. Mutat. 28, 578588.
  • Huijbers, I.J., Krimpenfort, P., Chomez, P., van der Valk, M.A., Song, J.Y., Inderberg-Suso, E.M., Schmitt-Verhulst, A.M., Berns, A., and Van den Eynde, B.J. (2006). An inducible mouse model of melanoma expressing a defined tumor antigen. Cancer Res. 66, 32783286.
  • Jacks, T., Remington, L., Williams, B.O., Schmitt, E.M., Halachmi, S., Bronson, R.T., and Weinberg, R.A. (1994). Tumor spectrum analysis of p53 mutant mice. Curr. Biol. 4, 17.
  • Kamijo, T., Zindy, F., Roussel, M.F. et al. (1997). Tumor suppression at the mouse INK4a locus mediated by the alternative reading frame product p19ARF. Cell 91, 649659.
  • Kannan, K., Sharpless, N.E., Xu, J., O’Hagan, R.C., Bosenberg, M., and Chin, L. (2003). Components of the Rb pathway are critical targets of UV mutagenesis in a murine melanoma model. Proc. Natl Acad. Sci. USA 100, 12211225.
  • Khlgatian, M.K., Hadshiew, I.M., Asawanonda, P., Yaar, M., Eller, M.S., Fujita, M., Norris, D.A., and Gilchrest, B.A. (2002). Tyrosinase gene expression is regulated by p53. J. Invest. Dermatol. 118, 126132.
  • Kumasaka, M.Y., Yajima, I., Hossain, K., Lida, M., Tsuzuki, T., Ohno, T., Takahashi, M., Yanagisawa, M., and Kato, M. (2010). A novel mouse model for de novo Melanoma. Cancer Res. 70, 2429.
  • Maldonado, J.L., Fridlyand, J., Patel, H. et al. (2003). Determinants of BRAF mutations in primary melanomas. J. Natl Cancer Inst. 95, 18781890.
  • Medic, S., and Ziman, M. (2010). PAX3 expression in normal skin melanocytes and melanocytic lesions (naevi and melanomas). PLoS ONE 5, e9977.
  • Molven, A., Grimstvedt, M.B., Steine, S.J., Harland, M., Avril, M.F., Hayward, N.K. et al. (2005). A large Norwegian family with inherited malignant melanoma, multiple atypical nevi, and CDK4 mutation. Genes Chromosomes Cancer 44, 1018.
  • Murase, D., Hachiya, A., Amano, Y., Ohuchi, A., Kitahara, T., and Takema, Y.J. (2008). The essential role of p53 in hyperpigmentation of the skin via regulation of paracrine melanogenic cytokine receptor signaling. Biol. Chem. 284, 43434353.
  • O’Hagan, R.C., Brennan, C.W., Strahs, A., Zhang, X., Kannan, K., Donovan, M., Cauwels, C., Sharpless, N.E., Wong, W.H., and Chin, L. (2003). Array comparative genome hybridization for tumor classification and gene discovery in mouse models of malignant melanoma. Cancer Res. 63, 53525356.
  • Pjanova, D., Molven, A., Akslen, L.A., Engele, L., Streinerte, B., Azarjana, K., and Heisele, O. (2009). Identification of a CDK4 R24H mutation-positive melanoma family by analysis of early-onset melanoma patients in Latvia. Melanoma Res. 19, 119122.
  • Pollock, P.M., Harper, U.L., Hansen, K.S. et al. (2003). High frequency of BRAF mutations in nevi. Nat. Genet. 33, 1920.
  • Puig, S., Ruiz, A., Castel, T., Volpini, V., Malvehy, J., Cardellach, F., Lynch, M., Mascaro, J.M., and Estivill, X. (1997). Inherited susceptibility to several cancers but absence of linkage between dysplastic nevus syndrome and CDKN2A in a melanoma family with a mutation in the CDKN2A (P16INK4A) gene. Hum. Genet. 101, 359364.
  • Randerson-Moor, J. et al. (2001). A germline deletion of p14(ARF) but not CDKN2A in a melanoma-neural system tumour syndrome family. Hum. Mol. Genet. 10, 5562.
  • Rane, S.G., Dubus, P., Mettus, R.V. et al. (1999). oss of Cdk4 expression causes insulin-deficient diabetes and Cdk4 activation results in beta-islet cell hyperplasia. Nat. Genet. 22, 4452.
  • Rizos, H., Puig, S., Badenas, C., Malvehy, J., Darmanian, A.P., Jiminez, L., Mila, M., and Kefford, R.F. (2001). Mutations in the INK4a/ARF melanoma susceptibility locus functionally impair p14ARF. J. Biol. Chem. 276, 4142441434.
  • Sarkar-Agrawal, P., Vergilis, I., Sharpless, N.E., DePinho, R.A., and Runger, T.M. (2004). Impaired processing of DNA photoproducts and ultraviolet hypermutability with loss of p16INK4a or p19ARF. J. Natl Cancer Inst. 96, 17901793.
  • van Schanke, A., van Venrooij, G.M., Jongsma, M.J., Banus, H.A., Mullenders, L.H., van Kranen, H.J., and de Gruijl, F.R. (2006). Induction of nevi and skin tumors in Ink4a/Arf Xpa knockout mice by neonatal, intermittent, or chronic UVB exposures. Cancer Res. 66, 26082615.
  • Serrano, M., Lee, H., Chin, L., Cordon-Cardo, C., Beach, D., and DePinho, R.A. (1996). Role of the INK4a locus in tumor suppression and cell mortality. Cell 85, 2737.
  • Sherr, C.J. (2006). Divorcing ARF and p53: an unsettled case. Nat. Rev. Cancer 6, 663673.
  • de Snoo, F.A., Hottenga, J.J., Gillanders, E.M. et al. (2008). Genome-wide linkage scan for atypical nevi in p16-Leiden melanoma families. Eur. J. Hum. Genet. 16, 11351141.
  • Sotillo, R., García, J.F., Ortega, S., Martin, J., Dubus, P., Barbacid, M., and Malumbres, M. (2001). Invasive melanoma in Cdk4-targeted mice. Proc. Natl Acad. Sci. USA 98, 1331213317.
  • Sturm, R.A., Duffy, D.L., Zhao, Z.Z., Leite, F.P., Stark, M.S., Hayward, N.K., Martin, N.G., and Montgomery, G.W. (2008). A single SNP in an evolutionary conserved region within intron 86 of the HERC2 gene determines human blue-brown eye color. Am. J. Hum. Genet. 82, 424431.
  • Walker, G.J., and Hayward, N.K. (2002). Pathways to melanoma development: lessons from the mouse. J. Invest. Dermatol. 119, 783792.
  • Walker, G.J., Kimlin, M.G., Hacker, E., Ravishankar, S., Muller, H.K., Beermann, F., and Hayward, N.K. (2009). Murine neonatal melanocytes exhibit a heightened proliferative response to ultraviolet radiation and migrate to the epidermal basal layer. J. Invest. Dermatol. 129, 184193.
  • Yang, G., Curley, D., Bosenberg, M.W., and Tsao, H. (2007). Loss of xeroderma pigmentosum C (Xpc) enhances melanoma photocarcinogenesis in Ink4a-Arf-deficient mice. Cancer Res. 67, 56495657.
  • Youssef, K.K., Van Keymeulen, A., Lapouge, G., Beck, B., Michaux, C., Achouri, Y., Sotiropoulou, P.A., and Blanpain, C. (2010). Identification of the cell lineage at the origin of basal cell carcinoma. Nat. Cell Biol. 12, 299305.
  • Zalaudek, I., Marghoob, A.A., Scope, A., Leinweber, B., Ferrara, G., Hofmann-Wellenhof, R., Pellacani, G., Soyer, H.P., and Argenziano, G. (2008). Three roots of melanoma. Arch. Dermatol. 144, 13751379.
  • Zhang, Y., Xiong, Y., and Yarbrough, W.G. (1998). ARF promotes MDM2 degradation and stabilizes p53: ARF-INK4a locus deletion impairs both the Rb and p53 tumor suppression pathways. Cell 92, 725734.
  • Zuo, L., Weger, J., Yang, Q. et al. (1996). Germline mutations in the p16INK4a binding domain of CDK4 in familial melanoma. Nat. Genet. 12, 9799.

Supporting Information

  1. Top of page
  2. Summary
  3. Significance
  4. Introduction
  5. Methods
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References
  10. Supporting Information

Figure S1. Kaplan–Meier curve for MM-free survival comparing spontaneous MM development between groups.

Figure S2. Representative images of non-tumorous skin from all of the three genotypes. Animals exposed to UVR but had not yet developed tumours.

Figure S3. Kaplan–Meier curve for MM-free survival comparing deep dermal melanomas with those developing from naevi.

Figure S4. In each case at least 10 tumours from each genotype, and at least two fields per tumour, were assessed.

Table S1. Anatomical location of lesions between genotypes.

FilenameFormatSizeDescription
PCMR_752_sm_fs1_re.ppt165KSupporting info item
PCMR_752_sm_fs2_re.pptx2094KSupporting info item
PCMR_752_sm_fs3_re.pptx92KSupporting info item
PCMR_752_sm_fs4_re.pptx105KSupporting info item
PCMR_752_sm_tableS1.doc77KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.