Abstract 2030: A single-cell spatially resolved map of colorectal cancer identifies novel spatial relationships between cancer cells and the microenvironment

Abstract

Understanding the tumor microenvironment (TME) requires more than just a catalog of cell types and gene programs. It is critical to see the spatial organization of the cells are and where they form multicellular interaction networks. Here we present a single-cell spatially resolved transcriptomic analysis of human mismatch repair deficient (MMRd) and proficient (MMRp) colorectal cancer (CRC) specimens. High tumor mutational burden MMRd tumors are known to have an immune response characterized by higher cytolytic T cell infiltrates compared to MMRp tumors, making them an ideal system for spatial single-cell profiling and understanding how the immune-driven programs differ between these tumors. MERFISH is a massively multiplexed single molecule imaging technology which can simultaneously capture and measure the quantity and distribution of hundreds to thousands of RNA species within single cells across a tissue1. We designed a MERFISH library of over 450 genes including genes important to proliferation, apoptosis, immune signaling, immune cell type pathways and other critical pathways in CRC. Patient samples, obtained commercially or through MGH, were hybridized with the designed MERFISH library and stained with a cell boundary marker to delineate cells across the tissue. We performed unsupervised clustering to identify cell types and we calculated spatial statistics to characterize how the cell type distribution varied between MMRd and MMRp tumors. We identified the cellular composition of each tumor, including immune and stromal cells, and the spatial distribution of these cell types. We were able to readily identify all cell types and states previously discovered by single-cell RNA sequencing2 in intact patient specimens, thus providing an accurate map of the cellular composition and spatial organization of these cells in the tumor microenvironment. We transformed these cell types into neighborhoods and discovered a highly organized spatial distribution of most cell types throughout the tumor. While spatial organization was observed in both MMRp and MMRd CRC we saw a significant shift in spatial organization between these tumor classifications notably in the immune population. Further, previously predicted multicellular interaction networks2 appeared as spatially organized structures in the tissue and were distinct in MMRd versus MMRp tumor specimens. Our data provide a richness of concrete hypotheses about which cells are working together, how these cells function cooperatively, and where these cells are located which will be critical in advancing therapy in these immunologically distinct types of colorectal cancer. These cancer maps are critical to truly understand the biology of CRC as well as identify avenues for the development of future therapies for CRC patients. Citation Format: Colles Price, Jonathan H. Chen, Karin Pelka, Sherry Chao, Michael Therrien, Timothy Wiggin, Nicolas Fernandez, Jiang He, George Emanuel, Genevieve Boland, Nir Hacohen. A single-cell spatially resolved map of colorectal cancer identifies novel spatial relationships between cancer cells and the microenvironment [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 2030.

Application of Artificial Intelligence to In Vitro Tumor Modeling and Characterization of the Tumor Microenvironment

“…These findings have the potential to lead to the development of treatment strategies that are specific to each tumor subtype. [ 85 ]…”

“…These findings have the potential to lead to the development of treatment strategies that are specific to each tumor subtype. [85] Building on these early ST techniques, NGS-based approaches, such as Visium's 10x Genomics spatial technology, boast an enhanced resolution of 55 microns, higher throughput, and increased sensitivity, allowing for genome-wide ST tissue profiling. [79,86,87] 10x Genomics' capacity to detect areas of abnormal gene expression allows for in-depth investigations into diseases associated with tissue function and structure and the identification of new biomarkers by linking gene expression data to the tissue footprint.…”

The promise and challenge of spatial omics in dissecting tumour microenvironment and the role of AI

“…Using a 450-gene MERFISH panel, Price et al. reported a shift in immune spatial organization between the two tumour subtypes i.e., human mismatch repair deficient and proficient tumors, opening new avenues for tumour subtype-specific treatment strategies ( 34 ).…”

“…MERFISH allows profiling of up to tens of thousands of RNA species at subcellular resolution, with a maximum 1 cm 2 capture area of FF/FFPE tissues. Using a 450gene MERFISH panel, Price et al reported a shift in immune spatial organization between the two tumour subtypes i.e., human mismatch repair deficient and proficient tumors, opening new avenues for tumour subtype-specific treatment strategies (34).…”