Abstract 3817: A single-cell, spatial multiomics atlas and cellular interactome of all major skin cancer types
Laura Grice,Guiyan Ni,Xinnan Jin,Minh Tran,Emily Killingbeck,Mark Gregory,Onkar Mulay,Siok-Min Teoh,Arutha Kulasinghe,Michael Leon,Sarah Murphy,Sarah Warren,Youngmi Kim,Quan Nguyen
DOI: https://doi.org/10.1158/1538-7445.am2022-3817
IF: 11.2
2022-06-16
Cancer Research
Abstract:Skin cancer is by far the most common cancer, encompassing squamous cell carcinoma (SCC), basal cell carcinoma (BCC), and melanoma. The diversity of cell types and tissue organisation in skin cancer remains poorly understood yet is required to improve diagnosis and treatment. In this work, we integrated six imaging and sequencing technologies to build the first spatial single cell reference for all three major skin cancer types and create a comprehensive skin cancer interactome. Using single-cell RNA-Seq (RNA) of >100,000 cells from 11 paired patient biopsies, we identified 28 SCC cell types, including 10 immune cell types, and found core suites of 39 cancer genes and 222 healthy genes shared across ≥80% patient samples. Using independent Nanostring Digital Spatial Profiling (RNA, protein), we validated most immune cell types and gene markers at protein and RNA levels. The enrichment of an immune signalling signature in SCC was further revealed by spatial Nanostring Single Molecular Imaging - SMI (RNA). Strikingly, we found the high consistency in mapping cell types in scRNAseq data and the independent SMI data, for example, the distribution of the three keratinocyte layers (basal, cycling and differentiated). This observation suggested the power of combining scRNAseq data with spatial SMI data. Furthermore, we implemented three approaches to validate the spatial distribution and cell type co-localisation by both Visium Spatial Transcriptomics (RNA), SMI (RNA) and Opal Multiplex Polaris (protein). Finally, cell-cell interactions were inferred at the global level using scRNAseq data (no spatial information) and Visium data (with spatial dimension), which were then validated at high throughput (517 ligands/receptors) and single-cell resolution using SMI. These in situ interaction maps were built across all three cancer types to create a comprehensive spatial interaction atlas of skin cancer. We also used targeted approaches with Polaris (protein) and RNAScope (RNA) to confirm and visualise clinically-important ligand-receptor pairs, including checkpoint inhibitor drug targets PD-1 and PD-L1. By integrating six distinct yet complementary spatial and single cell technologies, this study highlights the power of a spatial multi-omics approach for understanding cell types and their activities in cancer tissues. Citation Format: Laura Grice, Guiyan Ni, Xinnan Jin, Minh Tran, Emily Killingbeck, Mark Gregory, Onkar Mulay, Siok-Min Teoh, Arutha Kulasinghe, Michael Leon, Sarah Murphy, Sarah Warren, Youngmi Kim, Quan Nguyen. A single-cell, spatial multiomics atlas and cellular interactome of all major skin cancer types [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) nr 3817.
oncology