PS02.241: SINGLE CELL RNA SEQUENCING REVEALS THE HETEROGENEITY OF NON-CANCER CELL POPULATIONS WITHIN THE ECOSYSTEM OF ESOPHAGEAL ADENOCARCINOMA

Abstract

Abstract Background Esophageal adenocarcinoma (EAC) develops in a complex ecosystem that defines tumour evolution, response to treatment and patient outcomes. We have shown that high levels of cancer associated fibroblast (CAF) in the tumour microenvironment (TME) predicts poor outcome and is inversely related to tumour infiltrating lymphocyte (TILs) abundance. Bulk sequencing studies lack the resolution to dissect the phenotypic and functional heterogeneity and cell-cell interactions of the TME. We have applied single cell RNA sequencing to 12 EAC patients to address this challenge. Methods A total of 24 single-cell suspensions were prepared from resected specimens and paired normal tissue. Single cells and barcoded mRNA-binding micro-particles were combined in droplets containing cell lysis buffer using a custom-built microfluidic platform. Captured mRNA with a cell barcode and unique molecular identifier was reverse transcribed, amplified and sequenced. SEURAT (v2.1, R-package) was used to identify highly-variable genes and perform cell clustering. Results Analysis of 6859 of the highest quality cells using the 4167 most variable genes revealed 46 clusters which were divided into 12 broad populations. Antigen Presenting Cells (n = 189), B Cells (n = 265), Cancer Cells (n = 1449), Endothelial Cells (321), Fibroblasts (1690), Mast Cells (n = 184), Monocytes/Macrophages (n = 254), Plasma Cells (n = 208), Smooth Muscle, (n = 115), Squamous Epithelium (n = 751), T Cells (n = 1433). Analysis of publicly available bulk RNAseq datasets (TCGA) of EAC showed that tumours that were ‘hot’ for a CAF gene signature were ‘cold’ for a T-Cell signature. Subset analysis of the fibroblasts from tumour samples that were enriched for the same CAF signature revealed 3 subtly different clusters. One of these sub-populations differentially expressed genes associated with the Gene Ontology terms GO:00,40011 (locomotion) GO:0,006928 (movement of cell or subcellular component) and GO:00,48870 (cell motility). The two tumours with the highest ratio of this type of CAF to T-Cells were both found to have distant metastasis at resection. Conclusion EAC CAFs are a heterogeneous population with distinct biological functions which may have different implications for prognosis. These early results suggest that we are able to identify candidate biological processes that may describe the mechanisms through which CAFs and TILs influence outcome. Disclosure All authors have declared no conflicts of interest.