Abstract:
Exposure to solar ultraviolet (UV) light creates DNA damage that induces high levels of somatic mutations in
human skin cancers like melanoma. UV light most commonly causes cyclobutane pyrimidine dimers (CPDs)
and 6-4 photoproducts (6-4PPs) at dipyrimidine sequences (i.e., TT, TC, CT, and CC). Deamination of
cytidines in these lesions or mutagenic bypass leads to their frequent conversion to mutations. As a result, UV-
exposed cells and skin cancers are dominated by C to T and CC to TT substitutions in dipyrimidine sequences
that collectively constitute a UV mutation signature. Surprisingly, many driver mutations that contribute to
melanoma progression have sequence context and substitution characteristics that do not conform to the
canonical UV mutation signature. For example, the most common melanoma driver mutation, BRAF V600E,
involves a T to A substitution in a GTG sequence context. This difference in mutation characteristics between
most UV-induced mutations and melanoma driver mutations has led to the hypothesis that UV light induces
melanomagenesis by mechanisms other than the induction of mutations. However, we have recently provided
experimental evidence of mutations caused by rare, non-canonical UV lesions in whole genome sequenced
yeast and bioinformatics evidence of similar lesions in human clinical skin cancers. These lesions are likely
bulky photoproducts formed at TA, CA, and AC dinucleotides. Strikingly, the mutations associated with these
atypical UV photoproducts have identical characteristics to many recurrent driver mutations in melanoma,
suggesting that these rare lesions may play a significant role in causing skin cancer. The objective of this
proposal is to better define the characteristics of atypical UV photoproducts and their contribution to cancer
progression. In Aim I, we will utilize CPD and 6-4PP photolyases expressed in yeast to assign UV-induced
mutation classes to their corresponding lesion types as well as assess the ability of physiological UVB light to
induce mutation classes associated with atypical UV photoproducts in yeast and human cells. In Aim II, we will
analyze the genome-wide distribution of atypical TA photoproducts using a novel high throughput sequencing
method, called UVDE-seq. We will also characterize the formation of putative CA and AC photoproducts and
their contribution to oncogenic BRAF mutations. Finally, Aim III will identify DNA polymerases involved in the
error-free and error-prone bypass of TA and AC photoproducts. Successful completion of these aims will
provide new insights into the molecular causes of skin cancer, and thereby define a new paradigm of UV
mutagenesis that could potentially explain the epidemiological association of acute UV exposure with
increased melanoma incidence.
