High-throughput single-cell RNA-seq (scRNA-seq) is used to describe complex tissues by characterizing transcriptional states of individual cells. Defining a cell's position, both in regard to tissue margins and its social context, is essential for understanding the intrinsic and extrinsic variables that effect the transcriptional identity of individual cells. Conventional high-throughput scRNA-seq assays, however, decouple cells from their original locations within tissues. In situ hybridization readouts of gene expression and in situ sequencing preserve spatial information in tissues, but currently have a lower total read threshold than NextGen sequencing – imposing a restriction on either cell throughput or transcriptional breadth. Combining the above methods or using regional barcodes to define 2D positions have spatially-reconstructed tissue regions but are seldom employed in unspecialized laboratories. We describe SEgmentation by Exogenous Perfusion (SEEP), a rapid, integrated, method for providing 3D spatial-segmentation to scRNA-seq data. Tissues are divided into layers based on accessibility of a fluorescent dye allowing sorted cells to be characterized by transcriptomic and regional identity. We use SEEP to explore how the transcriptional states of cells in high-grade serous ovarian cancer vary with respect to intratumoral position. We describe an integrated method for correlating 3D radial-spatial cell positions with scRNA-seq data. By employing a basic stain-and-sort method using an off-the-shelf live-dead stain, we are able to correlate transcriptional profiles with radial-spatial positions in both radially symmetric tissues (e.g., spheroids, organoids, spherical tumor masses) and linear tissue samples (e.g., punch biopsies). SEEP uses an imaging-based calibration step to inform the parameters of a FACS and sequencing based measurement step. Across spheroid, organoid, and PDX models of high-grade serous ovarian cancer, we find common positional transcriptomic profiles, ranging from single-gene proclivities to multi-gene-set enrichments across tissue layers.

Citation Format: David B. Morse, Michele Ceribelli, Joachim De Jonghe, Aleksandra Michalowski, Christoph Muus, Maria Vias, Samantha Boyle, David A. Weitz, James Brenton, Jason D. Buenrostro, Craig Thomas, Tuomas P. Knowles. Spatially-segmented single-cell transcriptomics by diffusional accessibility to a small-molecule dye [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 5879.