Spatially Annotated Single Cell Sequencing for Unraveling Intratumor Heterogeneity

Intratumor heterogeneity is a major obstacle to effective cancer treatment. Current methods to study intratumor heterogeneity using single-cell RNA sequencing (scRNAseq) lack information on the spatial organization of cells. While state-of-the art spatial transcriptomics methods capture the spatial distribution, they either lack single cell resolution or have relatively low transcript counts. Here, we introduce spatially annotated single cell sequencing, based on the previously developed functional single cell sequencing (FUNseq) technique, to spatially profile tumor cells with deep scRNA-seq and single cell resolution. Using our approach, we profiled cells located at different distances from the center of a 2D epithelial cell mass. By profiling the cell patch in concentric bands of varying width, we showed that cells at the outermost edge of the patch responded strongest to their local microenvironment, behaved most invasively, and activated the process of epithelial-to-mesenchymal transition (EMT) to migrate to lowconfluence areas. We inferred cell-cell communication networks and demonstrated that cells in the outermost ~10 cell wide band, which we termed the invasive edge, induced similar phenotypic plasticity in neighboring regions. Applying FUNseq to spatially annotate and profile tumor cells enables deep characterization of tumor subpopulations, thereby unraveling the mechanistic basis for intratumor heterogeneity. scRNA-seq was performed in various regions at different distances from the center of a 2D epithelial tumor model using functional single-cell sequencing. Gene expression profiles in the center and at the edge of the tumor were analyzed to quantify the level of EMT in MCF10A cells. Two separate experimental setups were used: we either profiled the tumor model in concentric rings of equal width (i.e. center, middle and outer; approx. ~1000 um wide bands) or we profiled the invasive edge of the tumor model at a higher resolution (i.e. comparing the center with the outermost edge; ~10 cell wide band). Cells in the different regions are phototagged using photoactivatable Janelia Fluor 549 and photoactivatable Janelia Fluor 646 dyes, after which cells are sorted based on fluoresence and subjected to scRNA-seq. For each setup, 2 biological replicates are present.