Project summary / abstract The normal cell of origin (COO) from which a cancer arises is fundamental to our basic notions of cancer development. COOs are central to the standard of care for hematopoietic cancers, where they inform prognosis and guide therapy. However, COOs are not incorporated into the clinical paradigms for treating lung adenocarcinomas, an epithelial cancer that is among the leading causes of cancer death worldwide. The major reason for this is that we still don’t know the COO for most lung adenocarcinomas or whether the COO varies from patient to patient. One major scientific obstacle to addressing this question is the limited understanding of lung cell types, and epithelial types in general. A second obstacle is lineage infidelity that occurs during lung cancer progression, which can obscure the COO when one analyzes tumors morphologically or transcriptionally. As a result, the field has been required to rely on genetically engineered mouse models, which are still limited in their ability to fully recapitulate genotypes and clinical features of human lung adenocarcinomas. Recent work by us and others in the field has shown that genome-wide patterns of passenger mutations can provide a patient-specific COO signal. However, this hypothesis has not been tested rigorously for lung cancer or for other malignancies. In this study, we investigate how passenger mutation patterns inform COOs in lung cancer. In particular, we focus on noncoding surfactant protein gene (SPG) insertions and deletions, which our recent study (Cell 2017) and preliminary data establishes as one of the most common mutational events in lung adenocarcinoma (30-40% of patients). In this study, we hypothesize that this mysterious but highly prevalent mutational pattern is a somatic genetic “tattoo” of alveolar type II (AT2) cell origin in SPG mutant lung adenocarcinomas. We will investigate this through deep profiling of tumor adjacent tissue using targeted and single cell sequencing. In addition, we will investigate whether SPG mutant lung adenocarcinomas are associated with distinct evolutionary trajectories, with respect to their mutational, transcriptional, and histological state. Finally, we will build on our preliminary studies that establish compelling links between genome distributions of somatic single nucleotide variants (SNVs) in lung cancer and cell-type specific gene expression profiles obtained from single-cell RNA sequencing of benign lung. These preliminary data indicate that some lung adenocarcinomas may have proximal (club or basal cell) rather than distal (AT2 cell) origins. We will build on our initial findings, to generate and analyze higher depth atlases of healthy lung and correlate the patterns of cell type specific transcriptional and open chromatin profiles with genomic distributions of somatic SNVs. This work will provide some of the first direct evidence to map human lung adenocarcinoma to specific COOs. Furthermore, our establishment...