Part IV
Genetics & Molecular Drivers
Cutaneous melanoma is one of the most-sequenced cancers on Earth. Four genomic subtypes — BRAF, RAS, NF1, triple-WT — carve up the disease, joined by CDKN2A and TERT promoter mutations as near-universal cooperators. We finish at the structural basis of vemurafenib’s selectivity for BRAF V600E.
1. The TCGA Genomic Subtypes
The Cancer Genome Atlas Skin Cutaneous Melanoma study (TCGA, Cell 2015) sequenced 333 cutaneous melanomas and divided them into four mutually-exclusive driver subtypes by their dominant MAPK-pathway mutation:
| Subtype | Frequency | Hallmark mutations | Targeted Rx? |
|---|---|---|---|
| BRAF | ~50% | BRAF V600E (~80% of BRAF), V600K (~10%), other rare | Yes — BRAFi + MEKi |
| RAS | ~25% | NRAS Q61R/K/L; rarer KRAS, HRAS | No direct (MEKi alone modest) |
| NF1 | ~14% | NF1 LOF (truncations, large dels) | No direct (MEKi some activity) |
| Triple-WT | ~10% | KIT (acral/mucosal), GNAQ/GNA11 (uveal), other | Selective (KIT → imatinib) |
All four subtypes have hyperactive MAPK signalling as the convergent endpoint; they differ in which node of the pathway is broken. CDKN2A loss and TERT- promoter mutations cross all four subtypes as commonly co-occurring drivers of cell-cycle and replicative immortality.
2. BRAF V600E
BRAF encodes a serine/threonine kinase of the RAF family (ARAF, BRAF, CRAF) that sits downstream of RAS in the MAPK cascade. It is the single most common driver oncogene in cutaneous melanoma: ~50% of all cases harbour an activating BRAF mutation.
- V600E — valine-to-glutamate substitution at codon 600 within the activation segment of the kinase domain. ~80% of BRAF-mutant melanomas. Encoded by GTG→GAG — not a canonical UV signature; suggests a second mutational process (perhaps replication error in proliferating melanocytes) cooperating with UV.
- V600K — valine-to-lysine; ~10%; more common in older patients with chronically sun-damaged skin (high CSD).
- Rare V600 variants — V600R, V600M, V600D; functionally similar.
- Non-V600 — e.g. K601E, L597; activate BRAF as monomers or dimers; respond variably to BRAFi.
Mechanistically, V600E mutates the activation loop so that BRAF behaves as a constitutively active monomer with ~500× the basal kinase activity of WT BRAF. This is the structural feature exploited by vemurafenib (PLX4032) and dabrafenib: type-I½ ATP- competitive inhibitors that bind the DFG-in active conformation of monomeric mutant BRAF. They inhibit V600E ~100× more potently than they inhibit WT BRAF, sparing healthy cells.
BRAF V600E + Vemurafenib (PDB 3OG7)
Co-crystal structure of the BRAF kinase domain bearing V600E in complex with the type-I½ inhibitor vemurafenib (PLX4032). The ligand occupies the ATP-binding cleft and stabilises a DFG-in active conformation specific to mutant BRAF; this preferential binding to V600E underpins the ~100-fold therapeutic window over wild-type BRAF and the rapid tumour responses observed in BRIM-3 (Chapman et al., NEJM 2011).
3. The MAPK Cascade — the Common Pathway
All four melanoma genomic subtypes hyperactivate the RAS–RAF–MEK–ERK cascade. The healthy logic of the pathway is:
Activating mutations short-circuit this circuit at different nodes:
- RTK-driven (KIT in acral/mucosal): pathway driven from the top.
- RAS-driven (NRAS Q61): pathway driven below RTK; both BRAF and CRAF are recruited to the membrane and activated as dimers.
- RAF-driven (BRAF V600E): pathway driven as monomeric kinase; bypasses RAS.
- NF1-loss: NF1 is a RAS-GAP; loss elevates RAS-GTP indirectly.
The therapeutic implication: monomeric mutant BRAF is exquisitely targetable; dimeric BRAF/CRAF activity (RAS-driven, NF1-loss) is much harder to drug, which is why pan-RAF and second-generation type-II inhibitors are an active field. Combining BRAFi with MEKi (downstream node) blunts paradoxical pathway reactivation and roughly doubles progression-free survival vs BRAFi alone (COMBI-d, COMBI-v).
4. NRAS Q61 — the Second-Most Driver
NRAS mutations occur in ~25% of cutaneous melanomas and are mutually exclusive with BRAF mutations. Hot-spot codons:
- Q61R, Q61K, Q61L — the dominant melanoma NRAS variants. Disrupt the catalytic glutamine of the GTPase, locking NRAS in the GTP-bound active form.
- G12, G13 — rarer in melanoma than in other RAS-driven cancers.
NRAS-mutant melanoma has historically been undruggable directly. The first KRAS G12C inhibitors (sotorasib, adagrasib) do not address NRAS (different residue, different cancer). NRAS-mutant patients receive immune-checkpoint blockade, with MEK inhibitors as a modest direct-targeting option (binimetinib in NEMO trial, modest benefit). Worth noting: NRAS-mutant tumours appear to respond at least as well to anti-PD-1 as BRAF-mutants.
5. NF1-Loss Subtype
NF1 (neurofibromin 1) is a RAS-GAP — it accelerates GTP hydrolysis on RAS, returning it to its inactive state. Loss-of-function (~14% of cutaneous melanomas) disables this brake and elevates RAS-GTP indirectly.
- Mutations are typically truncations or large deletions, with biallelic loss.
- NF1-loss tumours have the highest TMB of any melanoma subtype (median ~30 mut/Mb).
- Patients tend to be older with chronically sun-damaged skin and head/neck primaries.
- Germline NF1 loss-of-function causes neurofibromatosis type 1 (von Recklinghausen disease) — a Mendelian syndrome with a modestly elevated melanoma risk among many other tumours.
- MEK inhibitors (selumetinib, mirdametinib) have activity in NF1-driven plexiform neurofibromas; their role in NF1-mutant melanoma is being tested.
6. KIT — the Acral & Mucosal Driver
KIT (CD117) is a receptor tyrosine kinase essential for melanocyte development — KIT-loss-of-function germline mutations cause piebaldism. Activating KIT mutations are uncommon in cutaneous melanoma overall (<3%), but enriched in:
- Acral melanoma (palms, soles, nail beds): KIT mutations in ~15%.
- Mucosal melanoma (oral, anorectal, vulvar, conjunctival): KIT in ~15–25%.
- Chronically sun-damaged head & neck melanomas: KIT in ~5%.
The hot-spot is exon 11 (juxtamembrane domain, especially L576P, V559D, W557R) and exon 13 (kinase domain, K642E). KIT-mutant patients can respond to imatinib ~25–30% objective response rates (Carvajal et al., JAMA 2011), with second-generation KIT inhibitors (nilotinib) salvaging some imatinib-resistant patients. Acral and mucosal melanomas are otherwise relatively checkpoint-resistant (lower TMB, lower SBS7 burden), so KIT-targeted therapy retains a place in this subgroup.
7. CDKN2A — Familial Melanoma & the Cell-Cycle Brake
The CDKN2A locus on 9p21 encodes two distinct tumour-suppressor proteins from alternative reading frames of overlapping exons:
- p16INK4A — binds CDK4/CDK6, blocking phosphorylation of RB and arresting G1→S progression.
- p14ARF (alternative reading frame) — binds and inhibits MDM2, stabilising p53.
Loss of CDKN2A therefore simultaneously disables the RB and p53 brakes — an elegantly parsimonious tumour-suppressor circuit that is broken in ~60% of cutaneous melanomas (somatic biallelic loss).
CDKN2A is also the major familial melanoma gene: germline loss-of-function mutations underlie the familial atypical multiple-mole melanoma (FAMMM) syndrome, in which carriers have:
- ~30–70% lifetime risk of cutaneous melanoma (vs ~2% population baseline).
- ~10–25% lifetime risk of pancreatic adenocarcinoma.
- Earlier onset (median ~30s) and frequent multiple primaries.
Other less-common melanoma-predisposition genes include CDK4 (gain-of- function R24C/H, <1% of families), BAP1 (uveal-cutaneous tumour predisposition), POT1 (telomere-maintenance), MITF E318K (modest risk), and MC1R R variants as a polygenic background.
8. MITF Amplifications & Phenotype Switching
MITF (microphthalmia-associated transcription factor) is the lineage transcription factor of melanocytes — the master regulator of pigmentation, survival, and proliferation. In melanoma:
- Amplifications occur in ~15–20% of metastatic melanomas; predict worse outcome.
- E318K — a germline variant affecting SUMOylation, recurrent in familial cases (~1.5–2× risk).
- Phenotype switching — reversible state changes between MITF-high (proliferative, pigmented, differentiated) and MITF-low (invasive, AXL-high, dedifferentiated) cells. Important driver of resistance to BRAF/MEK inhibition (MITF-low cells are intrinsically less MAPK-dependent).
Phenotype switching is a striking departure from a strict genetic model of cancer: a single melanoma cell can flip between transcriptional states under microenvironmental pressure (TGFβ, hypoxia, MAPK pressure) without any genetic mutation, and these states differ in drug sensitivity. This is one reason targeted-therapy resistance can develop within weeks — faster than any secondary mutation could fix.
9. TERT Promoter Mutations — the Commonest Non-Coding Driver
TERT encodes the catalytic subunit of telomerase — the enzyme that maintains telomeres and is repressed in most somatic cells. Two recurrent point mutations in the non-coding promoter of TERT — first reported by Horn et al. and Huang et al. back-to-back in Science 2013 — create new ETS-family transcription-factor binding sites and reactivate TERT transcription.
- C228T (chr5:1295228 G>A on the negative strand) — the dominant mutation; creates a CCGGAA ETS motif.
- C250T (chr5:1295250 G>A) — a second site creating a similar new motif.
- Combined frequency: ~70–80% of cutaneous melanomas; the single commonest non-coding driver mutation in cancer.
- Both mutations are canonical C→T at dipyrimidine — an SBS7/UV fingerprint, providing direct molecular evidence for UV causality at this locus.
- Co-occur with germline CDKN2A in some FAMMM kindreds; rare germline TERT promoter mutations cause familial melanoma.
Functionally, TERT promoter mutations rescue cells from replicative crisis as telomeres shorten — one of the canonical hallmarks of cancer (Cancer Hallmarks). They almost certainly arise early in melanoma evolution, often before CDKN2A loss, and are present in some benign nevi at low frequency.
10. Resistance & Cooperation Among Drivers
Melanoma genomes are not just lists of mutations — the order, cooperation, and evolutionary trajectory matter for therapy:
- Driver pyramid. A typical cutaneous melanoma carries: BRAF or NRAS or NF1 (one of); CDKN2A loss; TERT promoter mutation; PTEN loss in ~10%; TP53 loss-of-function in ~15%; ARID2 / IDH1 / RAC1 in subsets.
- Cooperation. BRAF V600E alone in melanocytes induces senescence (oncogene-induced senescence; OIS); cooperation with CDKN2A loss or MITF amplification is required to bypass OIS and enter clonal expansion.
- Resistance to BRAFi+MEKi. Mechanisms include NRAS Q61 acquisition (~20% of progressors), MEK1 P124 mutations, BRAF amplifications, BRAF splice variants, MITF amplification, AXL upregulation. ~50% of patients progress within 12 months.
- Resistance to anti-PD-1. JAK1/2 loss-of-function, B2M loss (loss of MHC-I), neoantigen loss, PTEN loss (which suppresses T-cell infiltration).
These mechanisms motivate the central therapeutic strategy of Part VII: combine checkpoint blockade (anti-PD-1 ± anti-CTLA-4) with targeted therapy in BRAF-mutant patients, sequence appropriately, and salvage with TIL therapy or oncolytic virus when the tumour escapes both. The genetics of melanoma has, for the first time in oncology, become a routine clinical decision-making tool at every visit.