Part III

UV Radiation & DNA Damage

The mutagen and the mutations. UVB excites pyrimidine bases directly to form cyclobutane and 6-4 photoproducts; UVA acts mostly through reactive oxygen intermediates. Both write a recognisable fingerprint — the SBS7 signature — on every cutaneous melanoma genome.

1. The Solar UV Spectrum at the Earth’s Surface

The sun emits a broad blackbody spectrum, but the atmosphere is a strong UV filter. At the Earth’s surface ultraviolet radiation is conventionally divided into three bands:

BandWavelengthSurface fluxSkin penetration
UVC100–280 nm~0% (absorbed by O₂/O₃)Stratum corneum
UVB280–320 nm~5% of solar UVEpidermis (esp. basal layer)
UVA320–400 nm~95% of solar UVEpidermis & superficial dermis
UVA1340–400 nmReaches melanocyte basement membrane
UVA2320–340 nm

The photon energy carried by each band is:

\( E = \dfrac{hc}{\lambda} \quad \Rightarrow \quad E_{\text{300 nm}} \approx 4.13 \text{ eV}, \quad E_{\text{360 nm}} \approx 3.45 \text{ eV} \)

Both bands carry enough energy to break covalent bonds in DNA, but their dominant mechanisms differ: UVB is absorbed directly by DNA bases (especially pyrimidines, with absorption peaking near 260 nm but with a long tail into UVB); UVA is absorbed mostly by endogenous photosensitisers (melanin, riboflavin, NADH, porphyrins) which transfer energy to molecular oxygen, generating reactive species.

2. UVB Photochemistry — Direct DNA Damage

When a UVB photon is absorbed by a pyrimidine base (thymine, cytosine, 5-methyl- cytosine), the base is promoted to an excited singlet state and can undergo [2+2] cycloaddition with an adjacent pyrimidine on the same strand (cis-syn). The result is a cyclobutane pyrimidine dimer (CPD): a covalent four-membered ring linking two adjacent pyrimidines.

  • Frequency: ~10⁵ CPDs per cell per minimal erythemal dose (MED) of sunlight.
  • Sequence preference: TT > TC ≈ CT > CC; biased toward 5'-pyrimidine-3'-pyrimidine runs.
  • Distortion: bends and unwinds DNA by ~7° — recognised by NER damage sensors.
  • Failure mode: across CPDs, DNA polymerase η preferentially inserts dAMP opposite each pyrimidine (the “A-rule”); if a CPD contains cytosine that has spontaneously deaminated to uracil, the result is C→T at that position.

This last point is critical: deamination of CPD-locked cytosine is several orders of magnitude faster than free cytosine, so CPDs at TC, CC and 5mCC dipyrimidines are biased mutation factories, not just damage sites. Almost the entire UV mutational signature traces to this single mechanism.

3. CPDs vs 6-4 Photoproducts

UVB also produces a second class of photoproduct, the pyrimidine-(6-4)-pyrimidone (6-4PP): a non-cyclobutane lesion in which the C6 of one pyrimidine is covalently linked to the C4 of the next. The two lesions differ:

PropertyCPD6-4 photoproduct
Yield per UVB photon~75%~25%
DNA distortion~7° bend~44° bend
NER repairSlow (~24 h half-life)Fast (~2 h half-life)
Block to replicationStrongVery strong
Major mutational outcomeC→T at dipyrimidineC→T (less common)

CPDs dominate the per-photon yield and are repaired more slowly — they are therefore the dominant mutational substrate in chronically UV-exposed tissue. Recent single-molecule studies (Premi et al., Science 2015) showed that “dark CPDs” can continue to form for hours afterUV exposure ends, driven by chemiexcitation from melanin radicals — which partly explains why melanocytes accumulate so much damage despite their pigment shield.

4. UVA Photochemistry — Oxidative Damage

UVA is much more abundant at the surface than UVB, and although less directly absorbed by DNA, it acts via two indirect mechanisms:

  • Type I photosensitisation. Excited photosensitiser (e.g. melanin radical, NADH) transfers an electron to or from DNA, generating base radicals.
  • Type II photosensitisation. Excited photosensitiser transfers energy to ground-state O₂, generating singlet oxygen ¹O₂.

Both pathways generate 8-oxoguanine as the dominant DNA lesion. 8-oxoG mispairs with adenine during replication, producing G→T transversions — a different mutational signature from UVB’s C→T at dipyrimidines. UVA also produces some CPDs (especially TT) via triplet-triplet energy transfer, but its dominant fingerprint is oxidative.

Why UVA matters for melanoma. UVA penetrates deeper than UVB, reaching the basal layer where melanocytes live. It is also the dominant emission of older tanning beds and accounts for the UV that passes through window glass and most clothing. Modern broad-spectrum sunscreens (with avobenzone, ecamsule, or zinc oxide) emphasise UVA blockade for this reason.

5. Nucleotide Excision Repair — the Defence Line

The dominant pathway for repair of CPDs and 6-4 photoproducts is nucleotide excision repair (NER): a multi-protein system that recognises bulky distortions of the helix, excises a ~24–32 nt damaged single-stranded patch, and re-synthesises from the undamaged template. Two sub-pathways:

  • Global-genome NER (GG-NER). XPC-RAD23B-CETN2 senses helical distortion anywhere in the genome.
  • Transcription-coupled NER (TC-NER). Stalled RNA Pol II at a lesion in a transcribed strand recruits CSA/CSB; preferentially repairs transcribed strand.
  • Both converge on the TFIIH-XPB-XPD helicase complex, XPA verification, XPG and ERCC1-XPF endonucleases (5' and 3' cuts), then DNA polymerase δ/ε gap-fill and ligation.

Xeroderma pigmentosum (XP) is the clinical demonstration of NER’s importance. Loss-of-function mutations in any of XPA-XPG, plus the variant XPV (DNA polymerase η), produce a syndrome of extreme photosensitivity, freckling from infancy, and a melanoma incidence ~1000× the general population — the highest known cancer-syndrome risk for any tumour. XP children have a median age at first skin malignancy of ~9 years.

6. Mutational Outcomes — Why C→T at Dipyrimidines?

The chain of events from UV photon to permanent mutation is:

  1. UVB photon → CPD or 6-4 PP at a dipyrimidine (TT, TC, CT, or CC).
  2. Within the dimer, cytosine deaminates to uracil (or 5mC to thymine) on a timescale of hours — vastly faster than the unconstrained free base.
  3. NER may or may not repair before replication.
  4. If NER fails, replication-bypass polymerase η encounters the lesion and inserts dA (the “A-rule”) opposite each pyrimidine.
  5. For original C (now U), inserting A creates a U:A pair that on the next round becomes T:A — a permanent C→T transition.
  6. If CC was the lesion, both cytosines may deaminate, producing the diagnostic CC→TT tandem mutation that is essentially pathognomonic of UV.

The genome-wide consequence: >75% of all somatic substitutions in cutaneous melanoma are C→T transitions at dipyrimidines, and CC→TT tandem mutations occur at a rate ~1,000× that expected from random C→T at non-dipyrimidine sites — a uniquely identifiable photochemical signature.

7. The COSMIC SBS7 (UV) Mutational Signature

The COSMIC catalogue of single-base substitution (SBS) signatures (Alexandrov et al., Nature 2020) decomposes whole-genome mutation profiles into recurring patterns generated by specific mutagenic processes. The UV signatures are SBS7a, SBS7b, SBS7c, SBS7d — now collectively just “SBS7 / UV”:

  • SBS7a, SBS7b. Dominated by C→T at TpC and CpC dipyrimidines — the classical CPD-deamination footprint.
  • SBS7c, SBS7d. Less abundant; distinct context preferences; thought to reflect 6-4PP repair and oxidative components.
  • SBS7 is the dominant signature in cutaneous melanoma, in BCC and SCC of sun-exposed sites, and in xeroderma pigmentosum tumours.
  • SBS7 is essentially absent from acral and mucosal melanomas, internal tumours, and uveal melanoma — the cleanest possible mutational evidence for UV causality of cutaneous melanoma.

Cutaneous melanomas carry one of the highest somatic mutation burdens in the cancer atlas: median ~16 mutations / Mb (approximately 50,000 somatic substitutions per genome), exceeded only by lung cancer in heavy smokers. This high burden is the reason melanoma carries an unusually rich neoantigen repertoire — the molecular basis of its sensitivity to checkpoint blockade.

Reference exposure metric. Tumour mutational burden (TMB) is reported per megabase. For melanoma the median is \( \sim 16 \text{ mut/Mb} \), vs. ~7–9 for NSCLC, ~3–5 for colorectal, and \( <1 \) for most pediatric cancers. TMB >10 mut/Mb is a tissue-agnostic FDA biomarker for pembrolizumab.

8. Dose-Response — Intermittent Intense, not Chronic

A central observation in melanoma epidemiology, separating it from squamous-cell carcinoma, is the intermittent-intense hypothesis:

  • SCC and BCC correlate roughly with cumulative lifetime UV: outdoor workers are at highest risk.
  • Cutaneous melanoma correlates more strongly with peak intermittent UV: severe sunburns, especially in childhood, beach holidays in normally pale individuals, indoor workers who get intense weekend exposure.
  • Outdoor workers actually have a slightly lower melanoma risk than indoor workers in some studies — the “outdoor-worker paradox”.

Mechanistically, intermittent intense exposure is hypothesised to overwhelm the MC1R/cAMP-mediated tanning response (which is slow), leaving the melanocyte under-protected during repeat exposures. Chronic exposure may instead cause squamous proliferation and enough cumulative pigmentation to be partly photoprotective.

The childhood window. Sunburn frequency before age 20 is among the strongest melanoma risk factors: ≥5 severe sunburns in adolescence approximately double lifetime risk (Whiteman et al., Cancer Causes Control 2001). The biological substrate is probably the long lifespan of pre-existing childhood melanocytes that accumulate UV mutations across the next 50+ years.

9. Tanning Beds & the IARC Reclassification

In 2009, the International Agency for Research on Cancer (IARC) reclassified ultraviolet-emitting tanning devices from Group 2A (probable) to Group 1 carcinogenic to humans — the same category as tobacco smoke and ionising radiation. The decision rested on:

  • A meta-analysis of 19 studies (Boniol et al., BMJ 2012): any tanning-bed use ~1.2× melanoma risk; first use before age 35 ~1.6×.
  • UVA flux from modern tanning beds is up to ~10–15× that of midday equatorial sun.
  • The IARC monograph found dose-response between number of sessions and risk.

Subsequent legislation (Brazil 2009: total ban; Australia 2014: total ban for commercial use; many US and EU jurisdictions: under-18 ban) is one of the few primary-prevention interventions in oncology with measurable population-level follow-through.

Sunscreens in evidence. Daily broad-spectrum SPF 15+ use was demonstrated to reduce melanoma incidence in the Australian Nambour trial 10-year follow-up (Green et al., J Clin Oncol 2011): ~50% reduction in primary melanoma. Modern SPF50 broad-spectrum formulations (containing avobenzone, octocrylene, tinosorb, and metallic-oxide UVA blockers) cover both UVB and UVA, the latter critically. Sunscreen is best used in addition to — not in place of — protective clothing, shade, and timing of outdoor activity.

From the genome (this part) we now turn to the specific oncogenes that this UV mutagenesis preferentially activates: BRAF, NRAS, NF1, KIT, CDKN2A, MITF, and the TERT promoter.

Key references for further reading. Pfeifer & Besaratinia, UV wavelength-dependent DNA damage and human melanoma, Photochem Photobiol Sci 2012; Premi et al., Chemiexcitation of melanin derivatives induces DNA photoproducts, Science 2015; Alexandrov et al., The repertoire of mutational signatures in human cancer, Nature 2020; Boniol et al., Cutaneous melanoma attributable to sunbed use, BMJ 2012; Green et al., Reduced melanoma after regular sunscreen use, J Clin Oncol 2011; Whiteman et al., Cancer Causes Control 2001 (childhood sunburns).
Share:XRedditLinkedIn