PROJECT SUMMARY: Cancer genome sequencing has identified thousands of somatic mutations in different types of human cancer. These mutations provide a valuable source of information as to the potential mutagenic events that have occurred and created the mutations, sometimes decades before the tumor develops. For certain cancers, convincing links between an exposure and cancer mutations have been established, best exemplified by typical sunlight-induced mutations (C to T mutations at dipyrimidine sequences) that are prevalent in melanoma and in non-melanoma skin cancers and by smoking- related G to T transversion mutations that are common in lung cancers from smokers but not nonsmokers. The types and frequencies of mutations in different tumor types is called the mutational signature. Several signatures also have a characteristic DNA strand bias related to, for example, the direction of transcription. In several human cancers, there are unusual and characteristic mutational signatures of unknown origin. These signatures are thought to arise from specific DNA damage, defective DNA repair or from endogenous processes. We hypothesize that we can identify and recreate mutagenic signatures using DNA damage mapping, a procedure in which we determine DNA lesions of a specific type at all sequence positions of the human genome, combined with mutational analysis, with the goal of assigning mutational signatures to etiological agents relevant for specific human cancers. One key challenge in this field is method development. The method needs to be sufficiently sensitive to detect rare DNA lesions and must be capable of doing so at all positions of the genome. We have developed such a method, which we named circle-damage-sequencing (CD-seq). We propose three Specific Aims. In the first Aim, we will use our new method to characterize and map reactive aldehyde-derived DNA adducts that we hypothesize play a crucial role in liver cancer. In the second Aim, we will test the hypothesis that 8-oxo-dG incorporated from the nucleotide pool is responsible for sequence-specific A to C transversion mutations observed as a dominant mutation type in human esophageal adenocarcinomas. The last Aim will have the goal of understanding the mechanisms of targeted mutagenesis by tobacco smoke carcinogens of the PAH class and how this targeting relates to lung and oral squamous cell carcinomas in tobacco users. Our DNA damage studies will be complemented by mutation assays to confirm that the relevant pathways induce the expected types of mutations. Our proposed work will provide mechanistic insights into the potential origin of human cancer mutations. We anticipate that our methods will aid in many future studies of DNA damage and repair and will help identify other mutagenic mechanisms.