SEARCH

SEARCH BY CITATION

Keywords:

  • anchorage-independent growth;
  • ERK;
  • invasion;
  • MAPK pathway;
  • melanocyte

Summary

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

The mitogen-activated protein kinase (MAPK) pathway is important in melanoma. In this pathway, DUSP6 phosphatase negatively controls the activation of extracellular signal-regulated (ERK) kinase. Through comparison of melanoma signalling pathways between immortal mouse melanocytes and their tumourigenic derivatives, retrieved from mouse xenografts, we identified a molecularly distinct subtype of melanoma, characterized by reduced ERK activity and increased DUSP6 expression. Overexpression of DUSP6 enhanced anchorage-independent growth and invasive ability of immortal mouse melanocytes, suggesting that increased DUSP6 expression contributes to melanoma formation in the mouse xenografts. In contrast, reduced tumourigenicity was observed after DUSP6 overexpression in human melanoma cells. A minority of thick human primary melanomas had high DUSP6 expression and the same poor melanoma-specific survival as the majority of thick primaries with low DUSP6 levels. We have demonstrated that DUSP6 is important in melanoma and that it plays a different role in our distinct subtype of mouse melanoma compared with that in classic human melanoma.


Significance

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

Melanoma is a highly aggressive tumour with a poor prognosis for patients with advanced disease because it is resistant to current therapies. Better understanding of the mechanisms underlying melanoma progression is needed to assist the development of new therapeutic interventions for this disease. To this end, we have identified a molecularly distinct melanoma subtype, characterized by low levels of mitogen-activated protein kinase (MAPK) pathway activation. Furthermore, we have demonstrated that DUSP6, which dephosphorylates extracellular signal-regulated (ERK) kinase, is an important player in melanoma and that it has an opposite role in our molecularly distinct melanoma subtype from that in classic human melanoma..

Introduction

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

Malignant melanoma is the most aggressive tumour of the skin. If melanoma is diagnosed early, it can be cured by surgical excision with an 80% success rate. However, metastatic malignant melanoma is resistant to current therapeutic approaches and spreads very quickly, with a median survival of only 6–9 months (Balch et al., 2001). Recent studies indicated that a combination of altered regulation of various effectors in different molecular pathways was involved in the progression of normal melanocytes to malignant melanoma (Smalley et al., 2006). A better understanding of the molecular changes underlying melanoma progression will contribute to the diagnosis, prognosis, classification and treatment of this disease.

The MAPK pathway has been reported to be constitutively active in melanoma (Meier et al., 2005). Up to 70% of human melanomas were characterized by mutations in BRAF, which constitutively activate the MAPK pathway. The most frequent BRAF mutation was a glutamate for valine substitution at position 600 (V600E) (Brose et al., 2002; Gorden et al., 2003). In addition, 5–36% of primary melanomas had activating mutations in NRAS (Carr and Mackie, 1994). However, whereas 93% of primary melanomas with the BRAF V600E mutation had activated ERK kinase (Uribe et al., 2006), overall only 54% of primary human melanomas showed ERK activation (Cohen et al., 2002; Uribe et al., 2006; Zhuang et al., 2005). Previous studies have also shown that some subtypes of melanoma without BRAF or NRAS mutations did not have high levels of ERK activation (Shields et al., 2007). Thus, the extent to which activation of ERK contributes to melanoma progression seems to depend on the melanoma subtype.

The discovery of most of the cysteine-dependent dual-specificity protein phosphatases (DUSPs) has occurred in the last 7 yr, initiating extensive interest in their role and regulation. The DUSP6 gene is a transcriptional target of the MAPK pathway (Bermudez et al., 2011). DUSP6, which dephosphorylates ERK kinase to act as a negative feedback regulator for the MAPK pathway, plays important roles in the maintenance of cellular homoeostasis in response to growth factors (Amit et al., 2007; Bermudez et al., 2010). Disruption of this feedback loop could cause neoplastic, and even malignant, transformation (Bermudez et al., 2010; Patterson et al., 2009). DUSP6 expression is often lost in pancreatic cancer because of promoter hypermethylation (Xu et al., 2005). Overexpression of DUSP6 led to suppression of cell growth and induced apoptosis in pancreatic cancer cell lines (Furukawa et al., 2003). DUSP6 levels are also reported to be low in lung and ovarian cancer cells, with restoration of DUSP6 levels suppressing cell growth (Chan et al., 2008; Okudela et al., 2009). There have been few reports investigating the importance of DUSP6 in melanoma progression.

In the present study, we investigated the role of DUSP6 in melanomagenesis and found that, while it suppressed tumourigenesis in a classic model of human melanoma, it stimulated tumourigenesis in a molecularly distinct melanoma subtype derived from immortal mouse melanocytes.

Results

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

Isolation of a new type of tumourigenic mouse melanocyte

Primary melanocytes were isolated from the skin of newborn albino mice as described (Selfridge et al., 2010). After infection with an SV40 T antigen-expressing retrovirus and clonal isolation, two immortalized melanocyte cell lines, 3-1-1 and 5-1, were derived from independent animals. Both showed characteristic melanocyte morphology and stained positively for melanocyte markers MITF, S100 and melan-A (Figure S1). To assess their in vivo tumourigenicity, both cell lines were xenografted into nude mice, together with human malignant melanoma cell line, A375, as a positive control (Figure 1). All 12 A375 xenografts grew very rapidly, taking only 21 days to reach the maximum permitted size. Compared with this, all 12 xenografts of mouse melanocyte line, 5-1, showed a long latent period before commencing rapid growth until all reached the maximum permitted size by 59 days. Xenografts of 3-1-1 showed an even longer latent period, before only 2 of 12 commenced rapid growth, reaching the maximum permitted size by 115 days. All resulting 5-1 and 3-1-1 xenografts showed the histological and immunohistochemical characteristics of amelanotic malignant melanoma (Figure S1D). The 5-1 T1 and T2 and 3-1-1 T1 and T2 cell lines were reisolated back into culture from independent xenografts. These tumourigenic cell lines retained their characteristic melanocyte morphology and melanocyte-specific markers. When 3-1-1 T1 cells were xenografted to mice again, all xenografts formed large tumours within 27 days, much more quickly than the parent 3-1-1 cell line and with a comparable growth rate to A375 human melanoma cells (Figure 1).

image

Figure 1.  Growth of mouse melanocyte and human melanoma xenografts. 107 cells were injected subcutaneously into both flanks of 8–12-week-old female nude mice, and the increase in mean tumour volume for each group was monitored twice weekly. There were six mice in each group. An animal was culled when a tumour reached 1 cm3. ♦, human melanoma A375; bsl00001, immortalized mouse melanocyte 5-1; bsl00066, immortalized mouse melanocyte 3-1-1; X, tumourigenic mouse melanocyte derivative 3-1-1 T1.

Download figure to PowerPoint

Although 3-1-1, 5-1 and their tumourigenic derivatives were maintained in medium supplemented with growth promoter TPA (12-O-tetradecanoylphorbol-13-acetate), all cell lines grew well in its absence with very similar cell cycle profiles, albeit rather more slowly (Table S1). All the cell lines retained p53 protein, indeed the level was higher than in mouse keratinocytes, but there was no evidence of UV-induced cell cycle arrest in G1, indicating that as expected, the p53 pathway was functionally inactivated by SV40 T antigen (Figure S2). Although the p53-induced cell cycle arrest pathway was inactive, all cell lines showed similar levels of UV-induced apoptosis, as determined by the presence of nuclei with sub-G1 DNA content and confirmed by Annexin V staining (Table S1).

These results suggest that introduction of SV40 T antigen into primary mouse melanocytes has produced an intermediate stage between the normal and malignant cell that requires additional genetic or epigenetic changes for full expression of the malignant phenotype. A similar conclusion was reached previously for melanocytes expressing the SV40 early region that were isolated from transgenic mice (Larue et al., 1993). This interpretation is supported by another report that SV40 T antigen-transformed human melanocytes did not undergo complete transformation to the malignant phenotype, remaining anchorage-dependent and unable to form xenografts (Zepter et al., 1995).

Altered signalling pathways in tumourigenic mouse melanocytes

To investigate the changes in our immortal melanocyte cell lines needed to form tumours in the xenograft experiment, protein lysates from immortal mouse melanocytes, their tumourigenic derivatives and the human A375 melanoma cell line were western-blotted to determine the levels of p-ERK1/2, p-MEK1/2 and DUSP6. Total ERK, MEK and GAPDH levels (to correct for variations in protein loading) were also determined. As expected, human melanoma cell line A375 showed high expression of p-ERK. To our surprise, expression of p-ERK was lower in all mouse tumour cell lines than their parental lines (Figure 2A). The expression of p-ERK1/2 in 3-1-1 tumour cell lines decreased 10-fold compared with 3-1-1, while expression in 5-1 tumour cell lines decreased 10- to 15-fold compared with 5-1 (P < 0.001, Figure 2B). The expression of p-MEK1/2 was also three- to fourfold lower in tumour cell lines compared with their parental lines (P < 0.001, Figure 2C, 2D). However, DUSP6 levels were higher in tumour cell lines (Figure 2E). The expression of DUSP6 in 3-1-1 tumour cell lines increased four- to sixfold compared with 3-1-1 and increased three- to fourfold in 5-1 tumour cell lines compared with 5-1 (P < 0.001, Figure 2F). Thus, in our melanocyte panel, tumourigenicity was correlated with increased DUSP6 expression and reduced MAPK pathway activation.

image

Figure 2.  Reduced p-ERK and p-MEK and increased DUSP6 levels in tumourigenic derivatives of immortalized mouse melanocytes. Protein levels are shown in mouse melanocytes 3-1-1 and 5-1, their tumourigenic derivatives 3-1-1 T1, 3-1-1 T2, 5-1 T1, 5-1 T2 and human melanoma A375 cells. (A) Western blots showing levels of phosphorylated ERK1 and 2 (p-ERK), total ERK1 and 2 (ERK) and loading control GAPDH. The positions and sizes (kDa) of the proteins are indicated. (B) Histogram showing mean level of p-ERK (±SEM) from three independent experiments. p-ERK levels are expressed relative to total ERK levels and are normalized to human melanoma A375 cells. (C) Western blots showing levels of p-MEK1/2, total MEK1/2 and GAPDH. (D) Histogram showing mean level of p-MEK (±SEM) from three independent experiments. (E) Western blots showing levels of DUSP6 and GAPDH. (F) Histogram showing mean level of DUSP6 (±SEM) from three independent experiments. DUSP6 levels are expressed relative to GAPDH and are normalized to human melanoma A375 cells. ***P < 0.001.

Download figure to PowerPoint

The PI3K-AKT pathway is also frequently activated in human melanoma (Cully et al., 2006; Hennessy et al., 2005). As AKT activation can result in the inactivation of RAF and reduced MAPK pathway activity (Palmieri et al., 2009; Zimmermann and Moelling, 1999), we also investigated p-AKT levels in our melanocyte panel. As with p-ERK, p-AKT levels were also lower in the tumour cell lines than in their parental lines (Figure S3A). Some distinct human melanoma subtypes with overexpression of c-kit and CDK4 do not require ERK activation (Smalley et al., 2008), but there was no evidence for altered CDK4 levels in the tumourigenic melanocyte cell lines, nor for changes in the β-catenin/Wnt signalling pathway (Figure S3D,E). The tumour suppressor proteins, p16 and PTEN, are frequently lost in melanoma, but both proteins were retained in the tumourigenic melanocyte cell lines (Figure S3B,C). Thus, reduced MAPK pathway activation was the most striking change identified in the tumourigenic melanocyte cell lines compared with their non-tumourigenic parent lines.

DUSP6 negatively regulates ERK in immortalized mouse melanocytes

To investigate the relationship between DUSP6 and ERK activation, non-tumourigenic 5-1 cells, which had low DUSP6 levels, were transfected with a DUSP6 expression plasmid that also contained a neomycin resistance marker. Cells stably expressing DUSP6 were selected with neomycin, and a pool of neomycin-resistant transfectant colonies was obtained. A pool containing >100 individual colonies was used, rather than individual colonies, to smooth out position effects associated with individual DUSP6 integrations and so obtain a more accurate estimate of the effects of DUSP6 overexpression. Levels of spontaneous or UV-induced apoptosis were unaffected by DUSP6 overexpression (Table S1). Protein lysates from 5-1, 5-1 DUSP6 transfection pool, 5-1 T1 and 5-1 T2 were western-blotted to detect the expression of DUSP6 and p-ERK. The DUSP6 expression level increased sixfold in the 5-1 DUSP6 transfection pool compared with non-transfected 5-1 cells, while the expression level of p-ERK was reduced 10-fold (Figure 3). We conclude, as expected, that DUSP6 negatively regulates p-ERK levels in immortalized mouse melanocytes.

image

Figure 3.  DUSP6 negatively regulates extracellular signal-regulated kinase (ERK) in immortalized mouse melanocytes. A DUSP6 expression construct was transfected into 5-1 cells. (A) Western blots showing levels of DUSP6, p-ERK, total ERK and GAPDH in 5-1, 5-1 DUSP6 transfection pool, 5-1 T1 and 5-1 T2 cells. (B) Histogram showing mean level of DUSP6 from two independent experiments. DUSP6 levels are expressed relative to GAPDH and are normalized to 5-1 cells. (C) Histogram showing mean level of p-ERK from two independent experiments. p-ERK levels are expressed relative to total ERK and are normalized to 5-1.

Download figure to PowerPoint

Increased DUSP6 levels stimulate tumourigenicity in immortalized mouse melanocytes

To determine the involvement of DUSP6 in melanoma progression of our immortal melanocyte cell lines, we compared anchorage-independent growth and invasive ability of the 5-1 DUSP6 transfection pool with non-transfected 5-1 cells and their tumourigenic derivatives. As expected, the average colony size in methylcellulose of the 5-1 tumour cell lines was four- to sixfold larger than for 5-1, and the colony number was 2.5-fold higher (P < 0.001, Figure 4A-C). The colony number and size in 5-1 cells overexpressing DUSP6 were increased to the levels seen in the tumourigenic 5-1 derivatives (Figure 4A). The average colony size of the 5-1 DUSP6 transfection pool was increased fivefold (Figure 4B), and the colony number was increased threefold (Figure 4C) compared with non-transfected 5-1 cells (P < 0.001).

image

Figure 4.  DUSP6 overexpression stimulates tumourigenicity and invasiveness of immortal mouse melanocytes. (A) Colony formation of 5-1, 5-1 DUSP6 transfection pool, 5-1 T1 and 5-1 T2 in methylcellulose. Images are representative of 5-1, 5-1 DUSP6 transfection pool and 5-1 T1 colonies at Day 12. (B) Histogram showing mean colony size (±SEM) in methylcellulose from three independent experiments. (C) Histogram showing mean colony number (±SEM) from three independent experiments. The total number of colonies per photographic field was counted. Each field is approximately 0.6 mm2 of the culture dish. At least 6 fields were counted for each dish. (D) Histogram showing the quantification of invasion. 5-1, 5-1 T1, 5-1 T2 and 5-1 DUSP6 transfection cells were seeded on transwell filters and allowed to invade through matrigel along a serum gradient. After 96 h, cells were labelled with calcein AM and visualized in the matrigel at 15-μm-depth intervals by Olympus FV1000 confocal microscopy. The relative cell number at 45 μm was calculated as the fluorescence intensity (positive pixels) at 45 μm expressed relative to the signal at 0 μm. **P < 0.01, ***P < 0.001.

Download figure to PowerPoint

5-1 Tumour cell lines were also more invasive than 5-1 through matrigel (Figure S4A). The cell number 45 μm into the matrigel, relative to the cell number at the origin, was 0.35 for the tumour lines, compared with only 0.002 for 5-1 (P < 0.001, Figure 4D). Chemotactic invasion in 5-1 cells overexpressing DUSP6 was also increased. The relative cell number of the 5-1 DUSP6 transfection pool 45 μm into matrigel was 0.15, 75-fold more than for non-transfected 5-1 (P < 0.01, Figure 4D).

We conclude that ectopic overexpression of DUSP6 in immortalized 5-1 melanocytes can increase tumourigenicity and invasiveness in the in vitro surrogate assays to the levels seen in the 5-1 tumourigenic derivatives isolated from xenografts that have increased DUSP6 levels compared with the immortalized 5-1 parent.

Constitutive activation of MEK increases p-ERK and DUSP6 levels in tumourigenic mouse melanocytes

Introduction of constitutively active MEK1/2 can result in increased phosphorylation of ERK (Seger et al., 1992). To investigate the functional significance of the very low levels of ERK activation in our tumour cell lines, we ectopically expressed constitutively active versions of MEK1 and MEK2 in 5-1 T2 cells. The MEK plasmids contained a hygromycin resistance marker, and, as with the DUSP6 transfection, a pool of resistant colonies was obtained. As expected, levels of p-MEK and p-ERK were dramatically increased in the 5-1 T2 MEK1/2 transfection pool compared with non-transfected 5-1 T2 cells and were now higher than in 5-1 cells (Figure 5A-C). In addition, the DUSP6 expression level increased almost twofold in the 5-1 T2 MEK1/2 transfection pool compared with non-transfected 5-1 T2 (Figure 5A,D).

image

Figure 5.  Constitutive activation of MEK increases phosphorylation of extracellular signal-regulated kinase (ERK) and DUSP6 expression in a tumourigenic derivative of immortalized mouse melanocytes. Constitutively active MEK1 and MEK2 expression constructs were transfected into 5-1 T2 cells. (A) Western blots showing levels of p-MEK, total MEK, p-ERK, total ERK and GAPDH in 5-1, 5-1 T1, 5-1 T2 and 5-1 T2 MEK1/2 transfection pool cells. (B) Histogram showing mean level of p-MEK from two independent experiments. p-MEK levels are expressed relative to total MEK and are normalized relative to 5-1. (C) Histogram showing mean level of p-ERK from two independent experiments. (D) Histogram showing mean level of DUSP6 from two independent experiments. DUSP6 levels are expressed relative to GAPDH and are normalized relative to 5-1 cells.

Download figure to PowerPoint

Constitutive activation of MEK reduces the invasive ability of tumourigenic mouse melanocytes

To determine the involvement of ERK in melanoma progression of our immortal mouse melanocytes, we assessed both anchorage-independent growth and invasive ability of the 5-1 T2 MEK1/2 transfection pool. Ectopic expression of constitutively active MEK resulted in a slight decrease in colony size and colony number, compared with 5-1 T2, but the difference was not significant (Figure 6A-C). However, expression of constitutively active MEK1/2 significantly suppressed the invasive capacity of 5-1 T2 cells (Figure S4B). The relative cell number 45 μm into the matrigel was 0.07 for the 5-1 T2 MEK1/2 transfection pool, threefold less than for non-transfected 5-1 T2 cells (P < 0.001, Figure 6D).

image

Figure 6.  Constitutive activation of MEK decreases invasiveness in a tumourigenic derivative of immortalized mouse melanocytes. (A) Colony formation of 5-1, 5-1 T1, 5-1 T2 and 5-1 T2 MEK1/2 transfection pool in methylcellulose. Images are representative of 5-1, 5-1 T2 and 5-1 T2 MEK1/2 transfection pool colonies at Day 7. (B) Histogram showing mean colony size (±SEM) in methylcellulose from three independent experiments. (C) Histogram showing mean colony number (±SEM) from three independent experiments. (D) Histogram showing the quantification of invasion through matrigel. The cell number at 45 μm into the matrigel is expressed relative to the cell number at 0 μm. **P < 0.01, ***P < 0.001.

Download figure to PowerPoint

We conclude that ectopic expression of constitutively active MEK in tumourigenic 5-1 T2 melanocytes can reduce invasiveness to approach the low level seen in the non-tumourigenic 5-1 parent line that has higher p-MEK and p-ERK levels than its tumourigenic derivatives.

Increased DUSP6 levels inhibit tumourigenicity of A375 human melanoma cells

In our distinct subtype of mouse melanoma, increased levels of DUSP6 result in tumourigenicity of our immortal mouse melanocytes. This is opposite to the role that has been reported for DUSP6 in some human cancers (Chan et al., 2008; Furukawa et al., 2003; Okudela et al., 2009). To investigate whether DUSP6 plays the same role in classic human melanoma, we used A375 human melanoma cells into which we had previously introduced the same DUSP6 expression vector as used here to overexpress DUSP6 in two independent pools of transfected colonies (Li and Melton, 2011). A375 DUSP6 transfection pools formed fewer and smaller colonies in methylcellulose than non-transfected A375 cells (Figure 7A-C). The average colony size and colony number of A375 DUSP6 transfection pools were both reduced twofold compared with non-transfected A375 cells (P < 0.01). Invasion was also significantly reduced in the A375 DUSP6-transfected pools (Figure S4C). The relative cell number of the A375 DUSP6 transfection pools 45 μm into the matrigel was decreased three- to fivefold compared with non-transfected A375 (P < 0.001, Figure 7D). We conclude that increased levels of DUSP6 reduce tumourigenicity and invasiveness in classic human melanoma, the opposite effect to that we have found in our molecularly distinct melanoma subtype.

image

Figure 7.  DUSP6 overexpression reduces tumourigenicity and invasiveness of human melanoma cells. A DUSP6 expression construct was previously transfected into A375 human melanoma cells resulting in DUSP6 overexpression (Li and Melton, 2011). (A) Colony formation of A375 and A375 DUSP6 transfection pools 3 and 4 in methylcellulose. Images are representative of colonies at Day 12. (B) Histogram showing mean colony size (±SEM) in methylcellulose from three independent experiments. (C) Histogram showing mean colony number (±SEM) from three independent experiments. (D) Histogram showing the quantification of invasion through matrigel. The cell number at 45 μm into the matrigel is expressed relative to the cell number at 0 μm. **P < 0.01, ***P < 0.001.

Download figure to PowerPoint

A subset of thick human primary melanomas with high levels of DUSP6 and poor melanoma-specific survival

To investigate further the relevance of our mouse melanoma subtype with high levels of DUSP6 to human melanoma, we stained a tissue microarray (TMA) of 44 thick (>2 mm) primary human melanomas for DUSP6 protein and quantified the immunohistochemical staining using the histoscore method. The frequency distribution of histoscores obtained is shown in Figure 8A. Overall, the staining was rather low (the median histoscore was between 100 and 125), but a subset of the samples had high histoscores. Representative cores with high and low histoscores are shown in Figure S5. Kaplan–Meier analysis showed that the quartile with the highest DUSP6 expression had a poor melanoma-specific survival profile that was indistinguishable from the combined group of the three low-expression quartiles (Figure 8B).

image

Figure 8.  High levels of DUSP6 in a subset of human melanomas with poor prognosis. 44 tissue cores from primary melanomas >2 mm thick were stained for DUSP6 by immunohistochemistry. (A) Frequency distribution of DUSP6 histoscores. (B) Kaplan–Meier plots for melanoma-specific survival are shown for the quartile of melanoma patients with the highest DUSP6 histoscores (150–300) and for the combined remaining three quartiles with lower histoscores (0–149). There was no significant difference in the survival curves by log rank test.

Download figure to PowerPoint

Discussion

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

Numerous studies have shown that activation of the MAPK pathway is involved in classic melanoma (Carr and Mackie, 1994; Meier et al., 2005; Smalley, 2003; Uribe et al., 2006). Oncogenic activation of RAS or BRAF frequently occurs in human melanoma, resulting in the activation of ERK and induction of various target genes (Dahl and Guldberg, 2007). However, previous studies have also identified subgroups of melanoma lacking mutations in NRAS and BRAF genes (Goel et al., 2006). A molecularly distinct melanoma subtype with overexpression of c-kit and CDK4 lacked ERK activation (Smalley et al., 2008). Furthermore, one study indicated that the absence of cytoplasmic ERK activation was an independent adverse prognostic factor in primary cutaneous melanoma and speculated that activation of some other pathways leading to tumour progression and adverse outcome may be important in melanoma (Jovanovic et al., 2008). Our tumourigenic mouse melanocyte cell lines have very low levels of ERK phosphorylation, but formed tumours very rapidly in a xenograft assay. These cell lines showed several other characteristics that were uncommon in classic melanoma: low expression of p-MEK and p-AKT, high expression of DUSP6 and retention of p16 and PTEN. These data suggest that these tumourigenic mouse melanocyte cell lines represent a molecularly distinct subtype of melanoma. It is possible that the genetic or epigenetic changes needed for tumourigenicity were already present at a very low frequency in the immortalized melanocytes xenografted into mice. Alternatively, these changes could have occurred after xenografting during the long lag before xenograft growth occurred. Either possibility would be compatible with the long delay before rapid xenograft growth of 3-1-1 cells, compared with the immediate rapid growth of 3-1-1 T1 xenografts.

DUSP6, which is itself induced by p-ERK, causes its inactivation by dephosphorylation. Disruption of this feedback loop is important in the initiation and development of pancreatic, lung and ovarian cancer (Bermudez et al., 2010). Our results showed that exogenous expression of DUSP6 dramatically decreased ERK phosphorylation and increased anchorage-independent growth and invasive ability of our immortal mouse melanocytes. This suggested that the increased levels of DUSP6 we observed in the tumourigenic derivatives of our immortal mouse melanocytes were necessary for their ability to grow as melanoma xenografts.

We next investigated whether the absence of activated ERK was required for tumourigenicity of our melanocytes in a xenograft assay. Ectopic expression of constitutively active MEK increased p-ERK levels and efficiently suppressed invasiveness of tumourigenic melanocytes in a transwell migration assay. This suggested that the inactivation of the MAPK pathway we observed in the tumourigenic derivatives of our immortal mouse melanocytes was necessary for their invasive ability. The accompanying reduction in anchorage-independent colony formation resulting from ectopic expression of constitutively active MEK in tumourigenic melanocytes was not significant. Constitutive MEK activation not only increased p-ERK levels, but also increased DUSP6 expression in our tumourigenic melanocytes. This agrees with a report that activation of the MAPK pathway can increase expression of DUSP6 through regulation of transcriptional factor ETS2 (Furukawa et al., 2008). The DUSP6 increase could contribute to anchorage-independent growth and so impair the suppression of colony formation caused by increased p-ERK levels. We also cannot exclude the possibility that DUSP6 controls the anchorage-independent growth of our tumourigenic melanocytes through a MAPK pathway-independent mechanism.

The human malignant melanoma cell line A375 contains the BRAF V600E mutation and has constitutively elevated ERK activity, which is required for its proliferation (Cheng et al., 2007). We demonstrated that ectopic expression of DUSP6 in A375 cells dephosphorylated ERK. Although the DUSP6-transfected A375 pools had the same growth rate as non-transfected A375 cells in routine cell culture (Li and Melton, 2011), they formed smaller and fewer anchorage-independent colonies and showed decreased invasive ability compared with non-transfected cells. Our data provide the first direct evidence for the role of DUSP6 in reducing tumourigenesis in classic human melanoma. Overexpression of DUSP6 has also previously been reported to reduce tumourigenesis in pancreatic (Furukawa et al., 2003), ovarian (Chan et al., 2008) and lung cancer (Okudela et al., 2009).

We have demonstrated that DUSP6 plays opposite roles in our tumourigenic mouse melanocytes compared with classic human malignant melanoma cells. Mice and humans may have different mechanisms for melanomagenesis, and DUSP6 may interact with different targets and control different pathways in human compared with mouse melanocytes. Alternatively, we have considered that our tumourigenic mouse melanocytes may resemble a distinct molecular subtype of human melanoma with decreased levels of p-ERK and p-AKT (Curtin et al., 2006; Shields et al., 2007). In addition, our mouse melanocytes did not contain the mouse equivalent of the BRAF V600E mutation. Some studies have shown that different mechanisms underlie melanoma progression in melanoma cell lines depending on whether they contain a BRAF mutation (Bloethner et al., 2005; Pratilas et al., 2009; Ryu et al., 2007).

We began an investigation of the possibility that the distinct molecular subtype of melanoma that we had identified from our mouse melanocyte work might be represented in human melanoma by searching the melanoma gene expression microarray data on the NCBI GEO database (http://www.ncbi.nlm.nih.gov/geoprofiles). In the largest data set of primary melanomas available (GDS1375), 10% of the primaries had high expression of DUSP6, while the remainder had low, or undetectable expression. Encouraged by this, we identified a similar subset of thick primary melanomas with high expression of DUSP6 protein. The subgroup of human primary melanomas with elevated DUSP6 expression showed a similar poor melanoma-specific survival profile to the thick primary group as a whole. This supports the relevance to human melanoma of the molecularly distinct melanoma subtype that we have identified from our work on immortalized mouse melanocytes.

In conclusion, there is very limited information on the role of DUSP6 in melanoma progression. Our studies provide the first direct evidence for the functions of DUSP6 in melanoma cells. Overexpression of DUSP6 stimulates melanoma formation in our distinct molecular subtype of melanoma, but acts to suppress tumourigenesis in classic human melanoma. We hope that our efforts will contribute to better understanding of the mechanisms underlying melanoma progression and assist the development of new therapeutic interventions for this disease.

Methods

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

Mammalian cell culture

Immortal mouse melanocyte cell lines 3-1-1 and 5-1 were clonally derived from independent primary melanocyte cultures of epidermis of newborn mice (Selfridge et al., 2010) by infection with a retrovirus expressing SV40 T antigen (Jat et al., 1986). The mouse stock was maintained on an outbred background, segregating for coat colour. 5-1 T1 and T2 and 3-1-1 T1 and T2 were cell lines reisolated back into culture from independent 5-1 and 3-1-1 xenografts. Mouse melanocyte cell cultures were maintained in RPMI-1640 medium supplemented with 10% foetal calf serum, 0.2 μM TPA, 25 U/ml penicillin and 25 μg/ml streptomycin. A375 human malignant melanoma cells were obtained from the European Collection of Cell Cultures and were maintained in Dulbecco’s modified Eagle’s medium, supplemented with 10% foetal calf serum, 25 U/ml penicillin and 25 μg/ml streptomycin. Plasmid pcDNA3.1/V5-His-DUSP6/MKP-3 was introduced into A375 cells as described (Li and Melton, 2011) and was introduced into 5-1 cells by electroporation, with selection for stable transformants in 2.4 mg/ml G418 (Invitrogen, Life Technologies Ltd, Paisley, UK). Plasmids pMCL-MKK1-R4F and pMCL-MKK2-KW71A express constitutively active forms of human MEK1 and MEK2 from the human cytomegalovirus promoter and also contain a hygromycin selectable marker (Mansour et al., 1994). These plasmids were introduced into 5-1 T2 cells by electroporation, with selection for stable transformants in 100 μg/ml hygromycin (Invitrogen). Selection was removed after pools of transfected colonies were isolated. Cell cycle distribution and levels of apoptosis were determined by flow cytometry of isolated propidium iodide-stained nuclei and Annexin V-stained live cells.

Western blotting

Protein extracts and western blotting were carried out as described by Selfridge et al. (2010). p-ERK, ERK, DUSP6, GAPDH, HRP-conjugated forms of goat anti-rabbit and rabbit anti-mouse secondary antibodies were used as described by Li and Melton (2011). p-MEK was detected using rabbit polyclonal antibody 9121 (1:500 dilution), and MEK was detected using rabbit polyclonal antibody 9122 (1:1000, both from Cell Signaling Technology, Inc., Danvers, MA, USA).

Methylcellulose assay

1.8% (w/v) agarose in phosphate-buffered saline was boiled, and the melted agarose solution was diluted 1:1 in fresh 2× Dulbecco’s modified Eagle’s medium supplemented with 10% foetal calf serum, 25 U/ml penicillin and 25 μg/ml streptomycin and then poured into a 6-well tissue culture plate (2 ml/well) and left until solid. 1.4% (w/v) methylcellulose solution was prepared in Dulbecco’s modified Eagle’s medium, supplemented with 10% foetal calf serum, 25 U/ml penicillin and 25 μg/ml streptomycin. Cells (1 × 104) were added to the 1.4% methylcellulose solution, and then, 2 ml was poured onto the bottom agar layer. The 6-well tissue culture plate was incubated at 37°C and 5% CO2. The colony number and size were determined and analysed by image j software (http://rsb.info.nih.gov/ij/).

Invasion assay

The transwell migration assay was carried out as described by Serrels et al. (2010). The plate was incubated at 37°C and 5% CO2 for 96 h, and then, cells were labelled with 5 μM calcein AM dye for 1 h and visualized in the matrigel at 15-μm intervals by confocal sectioning. The relative cell number in each section, determined from the fluorescence intensity, was analysed using image j software and expressed relative to the cell number in the section that represented the base of the transwell filter.

Xenograft

The mouse xenograft assay was carried out as described by Song et al. (2011).

Statistical analysis

Analysis was carried out using one-way anova with Bonferroni’s multiple comparison test or with Student’s t test.

Immunohistochemistry and histoscoring the TMA for DUSP6

Details of patients and tissue samples on the melanoma TMA have been described (Brown et al., 2011). Formalin-fixed and paraffin-embedded melanoma TMA sections were deparaffinized and rehydrated. Antigen retrieval was performed by incubation in 10 mM sodium citrate buffer (pH 6) for 40 min in a boiling water bath. TMA sections were then treated with 3% hydrogen peroxide for 10 min, followed by serum-free protein blocking reagent (DAKO, X0909, Dako UK Ltd, Ely, UK) for 1 h at room temperature and then incubated with primary antibody for DUSP6 (Abcam, Cambridge, UK, ab76310), diluted 1:500 in antibody diluent (DAKO, S0809), for 2 h at room temperature. Sections were then incubated with the EnVision™ Detection System (DAKO, K5007) as per the manufacturer’s instructions, counter-stained with haematoxylin, dehydrated and mounted. DUSP6 expression was quantified by histoscore. The histoscoring method for the quantification of immunohistochemical staining takes into account both the intensity of each stained cell [from 0 (no staining) to 3 (strong staining)] and the percentage of each category of staining in the sample to give a continuously variable score for 0–300.

Acknowledgements

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

We are grateful to Professor Toru Furukawa (Tohoku University School of Medicine, Japan), who kindly provided the DUSP6 plasmid and to Professor Natalie Ahn (University of Colorado, Boulder) who kindly provided the MEK plasmids. WL was supported by a China Scholarship Council/University of Edinburgh Scholarship. LS was supported by a grant (CZB/4/720) from the Scottish Chief Scientist Office. This work was supported by the Charon Fund.

References

  1. Top of page
  2. Summary
  3. Significance
  4. Introduction
  5. Results
  6. Discussion
  7. Methods
  8. Acknowledgements
  9. References
  10. Supporting Information
  • Amit, I., Citri, A., Shay, T. et al. (2007). A module of negative feedback regulators defines growth factor signaling. Nat. Genet. 39, 503512.
  • Balch, C.M., Buzaid, A.C., Soong, S.J. et al. (2001). Final version of the American Joint Committee on Cancer staging system for cutaneous melanoma. J. Clin. Oncol. 19, 36353648.
  • Bermudez, O., Pages, G., and Gimond, C. (2010). The dual-specificity MAP kinase phosphatases: critical roles in development and cancer. Am. J. Physiol. Cell Physiol. 299, C189C202.
  • Bermudez, O., Jouandin, P., Rottier, J., Bourcier, C., Pages, G., and Gimond, C. (2011). Post-transcriptional regulation of the DUSP6/MKP-3 phosphatase by MEK/ERK signaling and hypoxia. J. Cell. Physiol. 226, 276284.
  • Bloethner, S., Chen, B., Hemminki, K., Muller-Berghaus, J., Ugurel, S., Schadendorf, D., and Kumar, R. (2005). Effect of common B-RAF and N-RAS mutations on global gene expression in melanoma cell lines. Carcinogenesis 26, 12241232.
  • Brose, M.S., Volpe, P., Feldman, M. et al. (2002). BRAF and RAS mutations in human lung cancer and melanoma. Cancer Res. 62, 69977000.
  • Brown, E.R., Doig, T., Anderson, N., Brenn, T., Doherty, V., Xu, Y., Bartlett, J.M.S., Smyth, J.F., and Melton, D.W. (2011). Association of galectin-3 expression with melanoma progression and prognosis. European J. Cancer. doi:10.1016/j.ejca.2011.09.003.
  • Carr, J., and Mackie, R.M. (1994). Point mutations in the N-ras oncogene in malignant melanoma and congenital naevi. Br. J. Dermatol. 131, 7277.
  • Chan, D.W., Liu, V.W., Tsao, G.S., Yao, K.M., Furukawa, T., Chan, K.K., and Ngan, H.Y. (2008). Loss of MKP3 mediated by oxidative stress enhances tumorigenicity and chemoresistance of ovarian cancer cells. Carcinogenesis 29, 17421750.
  • Cheng, S.L., Huang-Liu, R., Sheu, J.N., Chen, S.T., Sinchaikul, S., and Tsay, G.J. (2007). Toxicogenomics of A375 human malignant melanoma cells. Pharmacogenomics 8, 10171036.
  • Cohen, C., Zavala-Pompa, A., Sequeira, J.H., Shoji, M., Sexton, D.G., Cotsonis, G., Cerimele, F., Govindarajan, B., Macaron, N., and Arbiser, J.L. (2002). Mitogen-actived protein kinase activation is an early event in melanoma progression. Clin. Cancer Res. 8, 37283733.
  • Cully, M., You, H., Levine, A.J., and Mak, T.W. (2006). Beyond PTEN mutations: the PI3K pathway as an integrator of multiple inputs during tumorigenesis. Nat. Rev. Cancer 6, 184192.
  • Curtin, J.A., Busam, K., Pinkel, D., and Bastian, B.C. (2006). Somatic activation of KIT in distinct subtypes of melanoma. J. Clin. Oncol. 24, 43404346.
  • Dahl, C., and Guldberg, P. (2007). The genome and epigenome of malignant melanoma. Apmis 115, 11611176.
  • Furukawa, T., Sunamura, M., Motoi, F., Matsuno, S., and Horii, A. (2003). Potential tumor suppressive pathway involving DUSP6/MKP-3 in pancreatic cancer. Am. J. Pathol. 162, 18071815.
  • Furukawa, T., Tanji, E., Xu, S., and Horii, A. (2008). Feedback regulation of DUSP6 transcription responding to MAPK1 via ETS2 in human cells. Biochem. Biophys. Res. Commun. 377, 317320.
  • Goel, V.K., Lazar, A.J., Warneke, C.L., Redston, M.S., and Haluska, F.G. (2006). Examination of mutations in BRAF, NRAS, and PTEN in primary cutaneous melanoma. J. Invest. Dermatol. 126, 154160.
  • Gorden, A., Osman, I., Gai, W., He, D., Huang, W., Davidson, A., Houghton, A.N., Busam, K., and Polsky, D. (2003). Analysis of BRAF and N-RAS mutations in metastatic melanoma tissues. Cancer Res. 63, 39553957.
  • Hennessy, B.T., Smith, D.L., Ram, P.T., Lu, Y., and Mills, G.B. (2005). Exploiting the PI3K/AKT pathway for cancer drug discovery. Nat. Rev. Drug Discov. 4, 9881004.
  • Jat, P.S., Cepko, C.L., Mulligan, R.C., and Sharp, P.A. (1986). Recombinant retroviruses encoding simian virus 40 large T antigen and polyomavirus large and middle T antigens. Mol. Cell. Biol. 6, 12041217.
  • Jovanovic, B., Krockel, D, Linden, D., Nilsson, B., Egyhazi, S., and Hansson, J. (2008). Lack of cytoplasmic ERK activation is an independent adverse prognostic factor in primary cutaneous melanoma. J. Invest. Dermatol. 128, 26962704.
  • Larue, L., Dougherty, N., Bradl, M., and Mintz, B. (1993). Melanocyte culture lines from Tyr-SV40E transgenic mice: models for the molecular genetic evolution of malignant melanoma. Oncogene 8, 523531.
  • Li, W., and Melton, D.W. (2011). Cisplatin regulates the MAPK kinase pathway to induce increased expression of DNA repair gene ERCC1 and increase melanoma chemoresistance. Oncogene. doi: 10.1038/onc.2011.426.
  • Mansour, S.J., Matten, W.T., Hermann, A.S., Candia, J.M., Rong, S., Fukasawa, K., Vane Woude, G.F., and Ahn, N.G. (1994). Transformation of mammalian cells by constitutively active MAP kinase. Science 265, 966970.
  • Meier, F., Schittek, B., Busch, S., Garbe, C., Smalley, K., Satyamoorthy, K., Li, G., and Merlyn, M. (2005). The RAS/RAF/MEK/ERK and PI3K/AKT signaling pathways present molecular targets for the effective treatment of advanced melanoma. Front. Biosci. 10, 29863001.
  • Okudela, K., Yazawa, T., Woo, T. et al. (2009). Down-regulation of DUSP6 expression in lung cancer: its mechanism and potential role in carcinogenesis. Am. J. Pathol. 175, 867881.
  • Palmieri, G., Capone, M., Ascierto, M.L., Gentilcore, G., Stroncek, D.F., Casula, M., Sini, M.C., Palla, M., Mozzillo, N., and Ascierto, P.A. (2009). Main roads to melanoma. J. Transl. Med. 7, 86.
  • Patterson, K.I., Brummer, T., O’Brien, P.M., and Daly, R.J. (2009). Dual-specificity phosphatases: critical regulators with diverse cellular targets. Biochem. J. 418, 475489.
  • Pratilas, C.A., Taylor, B.S., Ye, Q., Viale, A., Sander, C., Solit, D.B., and Rosen, N. (2009). (V600E)BRAF is associated with disabled feedback inhibition of RAF-MEK signaling and elevated transcriptional output of the pathway. Proc. Natl Acad. Sci. USA 106, 45194524.
  • Ryu, B., Kim, D.S., Deluca, A.M., and Alani, R.M. (2007). Comprehensive expression profiling of tumor cell lines identifies molecular signatures of melanoma progression. PLoS One 2, e594.
  • Seger, R., Ahn, N.G., Posada, J., Munar, E.S., Jensen, A.M., Cooper, J.A., Cobb, M.H., and Krebs, E.G. (1992). Purification and characterization of mitogen-activated protein kinase activator(s) from epidermal growth factor-stimulated A431 cells. J. Biol. Chem. 267, 1437314381.
  • Selfridge, J., Song, L., Brownstein, D.G., and Melton, D.W. (2010). Mice with DNA repair gene Ercc1 deficiency in a neural crest lineage are a model for late-onset Hirschsprung disease. DNA Repair 9, 653660.
  • Serrels, B., Sandilands, E., Serrels, A., Baillie, G., Houslay, M.D., Brunton, V.G., Canel, M., Machesky, L.M., Anderson, K.I., and Frame, M.C. (2010). A complex between FAK, RACK1, and PDE4D5 controls spreading initiation and cancer cell polarity. Curr. Biol. 20, 10861092.
  • Shields, J.M., Thomas, N.E., Cregger, M. et al. (2007). Lack of extracellular signal-regulated kinase mitogen-activated protein kinase signaling shows a new type of melanoma. Cancer Res. 67, 15021512.
  • Smalley, K.S. (2003). A pivotal role for ERK in the oncogenic behaviour of malignant melanoma? Int. J. Cancer 104, 527532.
  • Smalley, K.S., Haass, N.K., Brafford, P.A., Lioni, M., Flaherty, K.T., and Herlyn, M. (2006). Multiple signaling pathways must be targeted to overcome drug resistance in cell lines derived from melanoma metastases. Mol. Cancer Ther. 5, 11361144.
  • Smalley, K.S., Contractor, R., Nguyen, T.K. et al. (2008). Identification of a novel subgroup of melanomas with KIT/cyclin-dependent kinase-4 overexpression. Cancer Res. 68, 57435752.
  • Song, L., Ritchie, A.-M., McNeil, E.M., Li, W., and Melton, D.W. (2011). Identification of DNA repair gene Ercc1 as a novel target in melanoma. Pigment Cell Melanoma Res. 24, 966971.
  • Uribe, P., Andrade, L., and Gonzalez, S. (2006). Lack of association between BRAF mutation and MAPK ERK activation in melanocytic nevi. J. Invest. Dermatol. 126, 161166.
  • Xu, S., Furukawa, T., Kanai, N., Sunamura, M., and Horii, A. (2005). Abrogation of DUSP6 by hypermethylation in human pancreatic cancer. J. Hum. Genet. 50, 159167.
  • Zepter, K., Haffner, A.C., Trefzer, U., and Elmets, C.A. (1995). Reduced growth factor requirements and accelerated cell-cycle kinetics in adult human melanocytes transformed with SV40 large T antigen. J. Invest. Dermatol. 104, 755762.
  • Zhuang, L., Lee, C.S., Scolyer, R.A., McCarthy, S.W., Palmer, A.A., Zhang, X.D., Thompson, J.F., Bron, L.P., and Hersey, P. (2005). Activation of the extracellular signal regulated kinase (ERK) pathway in human melanoma. J. Clin. Pathol. 58, 11631169.
  • Zimmermann, S., and Moelling, K. (1999). Phosphorylation and regulation of Raf by Akt (protein kinase B). Science 286, 17411744.

Supporting Information

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

Figure S1. Immortalised mouse melanocytes and melanoma xenografts.

Figure S2. DNA damage response in immortalised mouse melanocytes and their tumourigenic derivatives.

Figure S3. Levels of cell signalling and cell cycle control proteins in immortalised mouse melanocyte cell lines and their tumourigenic derivatives.

Figure S4. Transwell migration assays. Cell migration at different depths through matrigel along a serum gradient. After 96 h migration, cells were labelled with calcein AM and visualized in the matrigel at 15 μm intervals by Olympus FV1000 confocal microscopy.

Figure S5. DUSP6 expression in human melanoma.

Table S1. Cell cycle distribution and apoptosis in immortalised mouse melanocytes.

FilenameFormatSizeDescription
PCMR_949_sm_fig-legends.pdf36KSupporting info item
PCMR_949_sm_fig1.tif5851KSupporting info item
PCMR_949_sm_fig2.tif820KSupporting info item
PCMR_949_sm_fig3.tif894KSupporting info item
PCMR_949_sm_fig4.tif2752KSupporting info item
PCMR_949_sm_fig5.tif4954KSupporting info item
PCMR_949_sm_table-1.pdf9KSupporting 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.