Unraveling heterogeneity in endometrial cancer via integrated single cell genotype-phenotype mapping

Endometrial carcinoma is a common but comprises a highly heterogeneous group of cancers, including subsets with hypermutation secondary to microsatellite instability or POLE mutation, and others with TP53-driven genomic instability. Defining molecular subtypes has led to improved prognostication and prediction of response to therapies. However, there remains a critical need to further refine classification, prognostication, and develop personalized treatment approaches. Current subtyping is based on bulk DNA profiling, which limits inferences of genetic subclones and phenotypic heterogeneity that emerge within individual cells. Further, comprehensive single cell genomics studies on endometrial cancer are sparse despite its increasing incidence worldwide. Here, we will develop and deploy novel single-cell technologies in all molecular subgroups of endometrial carcinoma. Though single cell technologies, including spatial platforms, are becoming more accessible, most technologies rely on access to fresh tissue samples for data generation. We propose to overcome these limitations via Genotyping In Fixed Transcriptomes (GIFT). GIFT uniquely combines genotyping mutations and profiling transcriptomes in the same single cells with full compatibility with FFPE tissue. This approach enables study of large, comprehensively characterized retrospective cohorts of endometrial carcinoma to resolve genotypic and phenotypic heterogeneity and relate genomic correlates to these rich phenotypic annotations, including tumor clonality patterns, evolution, and response to therapy. In Aim 1, we develop bioinformatics tools necessary to implement mutation (genotype) calling at the single cell and metacell levels, integrating these data with single cell transcriptomes to define subclones and construct lineage relationships. In Aim 2, we apply these computational tools in GIFT-profiling of existing cohorts of >600 endometrial carcinomas to map clonal trajectories at single timepoints and in longitudinal samples across many timepoints, thus delineating both static and dynamic features of heterogeneity, clonal states, and trajectories in endometrial cancer. In Aim 3, we resolve single cell spatial organization of endometrial cancer across genotypic subclones, transcriptional states, and surrounding microenvironment. Together, our work will provide a comprehensive profiling of endometrial carcinoma by refining the complex nature of endometrial cancer heterogeneity, clonal organization, and tumor evolution. We anticipate our results will directly inform the identification of biomarkers of treatment response and recurrence in endometrial carcinoma.