A single cell atlas of the human liver tumor microenvironment

• Published in: Molecular systems biology 2020-12, Vol.16 (12), p.e9682-n/a
• Main Authors: Massalha, Hassan, Bahar Halpern, Keren, Abu‐Gazala, Samir, Jana, Tamar, Massasa, Efi E, Moor, Andreas E, Buchauer, Lisa, Rozenberg, Milena, Pikarsky, Eli, Amit, Ido, Zamir, Gideon, Itzkovitz, Shalev
• Format: Article
• Language: English
• Subjects: Anatomy, Artistic; Animals; Atlases as Topic; Cholangiocarcinoma; EMBO03; EMBO22; Endothelial Cells - metabolism; Fibroblasts; Gene expression; Gene Expression Profiling; Gene Expression Regulation, Neoplastic; Gene Regulatory Networks; Gene sequencing; Hepatocytes; Hepatocytes - metabolism; human cell atlas; Humans; Liver; Liver cancer; Liver Neoplasms - genetics; Liver Neoplasms - immunology; Liver Neoplasms - pathology; Lymphocytes; Metastases; Metastasis; Mice; Patients; Ribonucleic acid; RNA; single cell RNAseq; Single-Cell Analysis; Smooth muscle; Spatial analysis; Spatial distribution; spatial transcriptomics; Stromal cells; Transcriptomics; Tumor cells; Tumor microenvironment; Tumor Microenvironment - genetics; Tumors; tumor‐stroma interactions; Zonation
• ISSN: 1744-4292; 1744-4292
• DOI: 10.15252/msb.20209682

Malignant cell growth is fueled by interactions between tumor cells and the stromal cells composing the tumor microenvironment. The human liver is a major site of tumors and metastases, but molecular identities and intercellular interactions of different cell types have not been resolved in these pathologies. Here, we apply single cell RNA‐sequencing and spatial analysis of malignant and adjacent non‐malignant liver tissues from five patients with cholangiocarcinoma or liver metastases. We find that stromal cells exhibit recurring, patient‐independent expression programs, and reconstruct a ligand–receptor map that highlights recurring tumor–stroma interactions. By combining transcriptomics of laser‐capture microdissected regions, we reconstruct a zonation atlas of hepatocytes in the non‐malignant sites and characterize the spatial distribution of each cell type across the tumor microenvironment. Our analysis provides a resource for understanding human liver malignancies and may expose potential points of interventions. SYNOPSIS Single cell transcriptomics and spatial methods are used to generate a cell atlas of the human liver tumor microenvironment, exposing recurring tumor‐stroma interactions and zonation patterns in the healthy and malignant tissue. A single cell atlas of the malignant and adjacent non‐malignant human liver is presented. Recurring stromal cell gene expression signatures are found in liver metastases and cholangiocarcinomas. Tumor and stromal cells communicate through a conserved ligand‐receptor interaction network. Spatial transcriptomics reveal zonated expression patterns in the malignant and non‐malignant liver. Graphical Abstract Single cell transcriptomics and spatial methods are used to generate a cell atlas of the human liver tumor microenvironment, exposing recurring tumor‐stroma interactions and zonation patterns in the healthy and malignant tissue.