The promising application of cell-cell interaction analysis in cancer from single-cell and spatial transcriptomics

Abstract

Cell-cell interactions instruct cell fate and function. These interactions are hijacked to promote cancer development. Single-cell transcriptomics and spatial transcriptomics have become powerful new tools for researchers to profile the transcriptional landscape of cancer at unparalleled genetic depth. In this review, we discuss the rapidly growing array of computational tools to infer cell-cell interactions from non-spatial single-cell RNA-sequencing and the limited but growing number of methods for spatial transcriptomics data. Downstream analyses of these computational tools and applications to cancer studies are highlighted. We finish by suggesting several directions for further extensions that anticipate the increasing availability of multi-omics cancer data.

Keywords: Cancer; Cell-cell interactions; Single-cell RNA-sequencing; Spatial transcriptomics.

Fig. 1.

The biological components of cell-cell interactions that can be inferred from single-cell and spatial data. A The different types of intercellular CCI, which include autocrine, paracrine, and juxtacrine signaling. Autocrine signaling occurs when a cell receives the same signals secreted by itself. Paracrine signaling occurs when a cell receives a signal secreted by nearby cells. Juxtacrine signaling occurs a cell receives a signal from a directly adjacent cell through physical contact. B Intracellular CCI is the downstream response to intercellular signaling that takes place inside cells. When a sufficient number of ligands bind to associated receptors, the receptors are activated. The signal received by receptors will be converted into an intracellular signal and activate transcription factors and target genes.

Fig. 2.

Common visualizations of CCI inference from cancer datasets A Circle plot showing directed interactions between cell types. B Chord diagram indicating outgoing interactions from different cell types. C Heatmap showing the total interaction strength between sender and receiver cell types. D Bubble plot displaying the interaction strength between cancer cells and other cell types with respect to different ligand-receptor pairs. L1, L2, L3 represent three different ligands, and R1, R2, R3, R4 represent four different receptors.