Analysis of genetic heterogeneity of tumor-infiltrating immune cells in human cancer by single-cell RNA sequencing

Abstract

Introduction: Tumor environment is established by various components including the malignant, stromal and immune cells. Single-cell transcriptome profiling of tumor samples allows the dissection of heterogeneous tumor cells and the neighboring stromal or immune cells. Precise characterization of the infiltrating immune cells may provide clues for novel immunotherapy strategies.

Methods: A total of 515 individual cells from 11 breast cancer patients and 162 cells from four advanced gastric cancer (AGC) patients were analyzed by single-cell RNA sequencing (RNA-seq). Reference single-cell transcriptomes for M1-type or M2-type macrophages were generated from normal blood-derived monocytes after in vitro differentiation.

Results: Copy number alteration patterns inferred from the single-cell RNA-seq data separated tumor cells and non-tumor cells. Most of the non-tumor cells were immune cells. In breast cancer*, three distinct immune cell clusters of T lymphocytes, B lymphocytes, and macrophages were identified. T lymphocytes displayed immunosuppressive characteristics with a regulatory or exhausted phenotype. B lymphocytes were divided into two subgroups, the anti-apoptotic naïve/memory cell group and the highly proliferative B cell group. In AGC, all of the detected immune cells were tumor-associated macrophages (TAMs). When compared to the reference transcriptomes of the M1 or M2 macrophages, an M2-biased tendency was observed in the AGC TAMs with a heterogeneous level of polarization. In comparison, TAMs originating from breast cancer or colorectal cancer showed both M1-biased and M2-biased cells.

Conclusions: This study demonstrates the power of single-cell RNA sequencing for the characterization of tumor-infiltrating immune cells to develop immunotherapeutic strategies. Single-cell transcriptome analysis subdivides the microenvironmental immune cells by cell types and their properties with heterogeneous levels of pathway activation. High resolution profiles can provide a new view on targeting the exhausted tumor-infiltrating lymphocytes or transforming the anti-inflammatory M2-type TAM to the pro-inflammatory M1-type.