Recent Progress 2015
The primary focus of our studies is melanoma, a cancer that affects over 12,000 people and causes over 2,000 premature deaths each year in the UK. Common cutaneous (skin) melanoma is the most common form of melanoma. It occurs on hair-bearing skin and is associated with exposure to the ultraviolet light that is present in sunlight and produced by sunbeds. The two most commonly mutated oncogenes in melanoma are NRAS (about 20% of cases) and BRAF (about 45% of cases). The proteins that these mutant genes produce activate a conserved signalling pathway that controls cell growth and survival and drives melanoma development. It was not known how activation of this pathway interacts with exposure to ultraviolet light to drive melanoma development, but drugs that inhibit the pathway can stop the growth of melanomas when NRAS or BRAF are mutated. However, although most patients respond to these treatments, they all eventually fail treatment and relapse with drug-resistant disease. Over the past year, we have studied how NRAS and BRAF interact with each other and with ultraviolet light to drive melanoma development. We have also continued to study drug resistance in order to improve our knowledge of melanoma biology so that new treatment strategies can be developed.
In 2014 we reported our investigations into the relationship between ultraviolet light and the NRAS/BRAF signalling pathway. We used a mouse model of melanoma in which expression of mutant BRAF in the melanocytes (the pigment cells in the skin that are the precursors of melanoma) induced melanoma in about half of the mice in about 12 months. When we exposed these mice to weekly low doses of ultraviolet light, they all developed melanoma within 7 months (Viros et al). Since ultraviolet light alone did not induce melanoma, these data show that even low doses of ultraviolet light that mimic weak sunburn in humans are sufficient to accelerate the development of melanoma in melanocytes that express mutant BRAF.
To investigate how ultraviolet light accelerates melanoma development, we sequenced the tumours and found that those from mice exposed to ultraviolet light presented a significantly higher number of mutations than those that had not been exposed. In particular, we observed a signature of DNA damage that is characteristic of DNA following exposure to ultraviolet light, suggesting that ultraviolet light directly damages melanocytes’ DNA. In agreement with this, we found that about 40% of the tumours that had been exposed to ultraviolet light carried mutations in the tumour suppressor TP53, whereas no TP53 mutations were seen in the tumours that had not been exposed. We validated this result by showing that mutant BRAF and mutant TP53 cooperate to drive melanoma development even without ultraviolet light, establishing that TP53 is a bona fide target of ultraviolet light in melanoma and that it cooperates with BRAF to drive this disease. Finally, we demonstrated that while sunscreen delays melanoma onset, it only provided partial protection. This is because although sunscreen is very efficient at absorbing ultraviolet light, a small amount escapes and damages the DNA. Sunscreen clearly plays an important role in sun protection, but it is important to note that it does not offer complete protection and so it should be used in combination with other sun avoidance strategies to provide full protection from the damaging effects of the sun’s rays.
Figure 1. Skin from animals expressing BRAFV600E in their melanocytes 7 days after exposure to UVR. Exposed skin (top) and protected skin (bottom).
Although epidemiological studies have implicated ultraviolet light in common cutaneous melanoma, the role of this carcinogen in other forms of melanoma is less clear. In 2014 we reported the genomic landscape of acral melanoma, a relatively rare disease that affects the non-hair bearing skin of palms, soles and nail beds. In line with its distinct epidemiological and clinical profiles and compared to common cutaneous melanoma, we found acral melanoma has a relatively low mutation burden. Moreover, we did not observe the UV DNA damage signature that is a feature of the genomes of common cutaneous melanoma and we observed a distinct spectrum of mutations. In line with epidemiological and chromosomal-level genomic studies, our findings show that acral and common cutaneous melanoma are distinct diseases that will likely need to be treated differently. Notably, the genomic landscape of acral melanoma is more similar to mucosal melanoma than to common cutaneous melanoma, and in line with these findings, mucosal melanoma is also not associated to exposure to ultraviolet light.
As described above, genomic approaches can provide insight into the biology and aetiology of melanoma. Another use of this technology is to investigate mechanisms of drug resistance. In 2014, we reported a case of a patient who did not respond to a BRAF drug despite the presence of a BRAF mutation. We performed whole genome sequencing of the patient’s tumours and this revealed that in addition to the BRAF mutation, the tumours also carried mutations in two genes called PTEN and GNAQ. Our functional studies demonstrated that the mutant proteins these genes encoded reduced the tumour’s response to the BRAF drug. Thus, in this patient the tumours were resistant before the treatment even started (intrinsic resistance). We posit that knowledge such as this will be critical to optimising treatment for patients by providing a platform of knowledge that can be used to individualise or personalise treatment to ensure the best outcomes for each patient.
Finally, although NRAS and BRAF mutations each occur in a high proportion of melanomas, they are rarely coincident, suggesting that they satisfy similar requirements to melanoma cells. In rare cases however, NRAS and BRAF mutations are coincident, but this always involves rare mutations that inactivate rather than activate BRAF. To study these rare forms of melanoma, we again used mouse models. We found that inactive mutants of BRAF cooperated with mutant NRAS and mutant KRAS to induce melanoma, (Pedersen; Sanchez-Laorden). We have previously reported that this cooperation occurs because if BRAF is inhibited in the presence of mutant RAS, it drives paradoxical activation of the pathway. This occurs because when drugs inactivate BRAF, it binds to and stimulates a closely related protein called CRAF, which then drives hyper-activation of the pathway. Importantly, we found that in addition to accelerating the development of RAS-driven tumours, paradoxical activation of the pathway stimulated melanoma dissemination (metastasis) in the mice. More importantly, we found that when melanoma develops resistance to BRAF inhibitors, the inhibitors mediate a similar re-activation of the pathway and this then stimulates tumour dissemination. Our data suggests therefore that not only do the tumours stop responding to the drugs when they develop resistance, but that they use the drugs to drive tumour dissemination through the body. Clearly, knowledge such as this is critical for optimising patient treatment and for knowing when to withdraw treatments that are no longer working.
Figure 2. Personalised medicine in melanoma. Early detection of relapse and elucidation of the mechanisms of resistance to therapies are needed to guide clinical care and personalise treatment decisions for individual patients. The Molecular Oncology group integrates new diagnostic tools using a combination of techniques to optimise personalised care and improve patient outcomes. These include proteomics, metabolomics, genetically engineered modified mice (GEMM), pathology, cell lines, whole exome sequencing (WES), patient derived xenografts (PDXs), and circulating tumour DNA (ctDNA).