Poster Presentation 43rd Lorne Genome Conference 2022

Building a spatial single-cell multi-omics atlas and cellular interactome for skin cancer (#236)

Laura Grice 1 2 , Guiyan Ni 1 , Xinnan Jin 1 , Minh Tran 1 , Onkar Mulay 1 , Siok Min Teoh 1 , Emily Killingbeck 3 , Mark Gregory 3 , Youngmi Kim 3 , Katharina Devitt 4 , Arutha Kulasinghe 4 , Sarah Warren 3 , Kiarash Khosrotehrani 4 , Mitchell Stark 4 , Quan Nguyen 1
  1. The University of Queensland, St Lucia, QLD, Australia
  2. School of Biomedical Sciences, The University of Queensland, St Lucia, QLD, Australia
  3. NanoString Technologies, Seattle , WA , USA
  4. Diamantina Institute, The University of Queensland, St Lucia, QLD, Australia

Introduction

Skin cancer is by far the most common cancer, encompassing squamous cell carcinoma (SCC), basal cell carcinoma (BCC), and melanoma. 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 three major skin cancer types. Furthermore, we investigated cell-cell interactions to create a comprehensive skin cancer interactome.

 

Methods and results

 

Using single cell RNA-Seq (RNA) of >50,000 cells from 11 paired patient biopsies, we identified core suites of 39 cancer genes and 222 healthy genes shared across ≥80% patient samples. The cancer suite contained known SCC biomarkers and displayed marked immune pathway enrichment compared to healthy samples. We identified 21 cell types, including 10 immune cell types. Most immune cell types and gene markers were validated at protein and RNA levels with Nanostring Digital Spatial Profiling (RNA, protein). The enrichment of an immune signalling signature in SCC was further revealed by Nanostring Single Molecular Imaging (RNA), where 517 ligand-receptor genes were measured at subcellular resolution within in situ spatial context. Spatial distribution and cell type co-localisation was mapped and validated using 10x Visium Spatial Transcriptomics (RNA). Finally, we used Opal Multiplex Polaris (protein) and RNAScope (RNA) to confirm and visualise clinically-important ligand-receptor pairs, including checkpoint inhibitor drug targets PD-1 and PD-L1.  

 

Conclusions

By integrating six distinct yet complementary spatial and single cell technologies, we have built a comprehensive single-cell and spatial atlas of skin cancer, spanning SCC, BCC, and melanoma. We also constructed a comprehensive interactome between skin cancer cell types in spatial context. This study highlights the power of a spatial multi-omics approach for understanding cell types and their activities in cancer tissues.