Defining signatures of epigenetic sensitization to lung cancer in a mouse model

Epigenetic perturbations in cancer are widely recognized but poorly understood. A detailed understanding of how latent epigenetic information contributes to tumorigenesis would enable us to stratify risk among vulnerable patients, classify tumors based on epigenetic state, and better target therapeutic approaches. Animal models of early-onset epigenetic perturbation offer a unique opportunity to study the potential contribution of epigenetic signatures to prediction of cancer risk, since epigenetic changes precede tumor initiation in these models and can therefore be separated from secondary effects of tumorigenesis. In preliminary studies, we found that epigenetic changes induced by loss of the epigenetic regulator KDM6A (UTX) in the paternal germ line in mice result in elevated lung tumor burden in offspring even when the mutant Kdm6a allele itself is not inherited, implying that these offspring carry silent epigenetic perturbations that predispose to cancer. These mice are an excellent experimental model for defining signatures of epigenetic sensitization that predict cancer risk. Our central hypothesis is that preexisting, detectable epigenetic perturbations in lung tissue interact with genetic driver mutations to enhance lung tumorigenesis, and that the resulting tumors have a distinct epigenetic and genetic signature. The objective of this study is to define a signature of these preexisting epigenetic changes in the lung in a mouse model and determine if a memory of this signature can be detected in lung tumors. Aim 1 will define the transcriptional and epigenetic phenotypes of normal lung tissue in epigenetically sensitized mice. Aim 2 will evaluate the molecular phenotypes of epigenetically sensitized lung tumors using both spontaneous tumors and tumors induced by activated KRAS. If both are successful, the results of the two Aims will be compared to identify signatures of epigenetic sensitization that are present in histologically normal tissue and that persist in tumors. We expect this work to reveal new markers for epigenetic risk of lung tumorigenesis and define the molecular signatures associated with prior epigenetic sensitization in tumors. Future work will determine the extent to which such signatures can be applied to human populations in the clinic.