High-resolution spatial transcriptomics through light patterning


Year of Award:
2020
Status:
Complete
Award Type:
R21
Project Number:
CA246358
RFA Number:
RFA-CA-19-019
Technology Track:
Molecular & Cellular Analysis Technologies
PI/Project Leader:
SEELIG, GEORG
Other PI or Project Leader:
WU, DAVID
Institution:
UNIVERSITY OF WASHINGTON

PROJECT SUMMARY The cellular composition of a tumor as well as the spatial arrangement of cells within the tumor are major determinants of the response to therapy and the emergence of resistance. To improve our understanding of tumor heterogeneity, accelerate the discovery of new drug targets or enable better patient stratification it is thus necessary to develop tools that can resolve molecularly defined cell types within a tumor and capture their spatial relationships. Driven by progress in single-cell RNA sequencing (scRNA-seq) technologies, a complete census of molecularly defined cell types within a tumor is now within reach. However, because scRNA-seq requires dissociated cells and cannot preserve information about the spatial arrangement of cells in their original context, it gives an incomplete picture of the relationship between gene expression, cell type identity and tumor architecture. The need for technologies that measure gene expression in single cells while retaining position information has long been recognized, but existing solutions have insufficient cellular throughput, spatial resolution, or gene detection sensitivity. We propose to develop Combinatorial Light-Activated Spatial Sequencing (CLASSeq), a transformative approach to spatial transcriptomics that overcomes these limitations. CLASSeq uses patterned light illumination to attach DNA barcodes encoding location information to all cells of interest within a tissue section, with spatial resolution limited only by the wavelength of light. Spatial barcodes are sequenced together with cellular transcriptomes after dissociating the tissue into individual cells or nuclei, and tissue-wide gene expression patterns are computationally recreated. Because sequencing is performed after dissociation, any established scRNA-seq workflow can be used, enabling high sensitivity and cell throughput. To achieve high throughput and reproducibility, and facilitate wide adoption, we will work toward automating the labeling workflow by constructing a prototype instrument that integrates fluidics for barcode delivery with patterned illumination. To validate our approach and demonstrate its utility to cancer research, CLASSeq will be used to characterize cellular diversity and organization in Hodgkin lymphoma, a mature B-cell lymphoma in which the tumor microenvironment niche is critical to the tumor's success for host immune evasion and thus governs the response or lack thereof to clinical immune modulatory therapies.