Understanding tumor ecosystems by single-cell sequencing: promises and limitations

Abstract

Cellular heterogeneity within and across tumors has been a major obstacle in understanding and treating cancer, and the complex heterogeneity is masked if bulk tumor tissues are used for analysis. The advent of rapidly developing single-cell sequencing technologies, which include methods related to single-cell genome, epigenome, transcriptome, and multi-omics sequencing, have been applied to cancer research and led to exciting new findings in the fields of cancer evolution, metastasis, resistance to therapy, and tumor microenvironment. In this review, we discuss recent advances and limitations of these new technologies and their potential applications in cancer studies.

Fig. 1

State of the art of single-cell sequencing technologies. Single-cell sequencing technologies have been designed for almost all the molecular layers of genetic information flow from DNA to RNA and proteins. For each molecular layer, multiple technologies have been developed, all of which have specific advantages and disadvantages. Single-cell multi-omic technologies are close to comprehensively depicting the state of the same cells. We apologize for the exclusion of many single-cell sequencing methods due to the limited figure space

Fig. 2

Spatial heterogeneity of tumors. A tumor is a complex ecosystem composed of various cell types which show heterogeneous spatial distributions. The cell types within a tumor generally contain cancer cell clones, normal cells that have not been transformed, stromal cells, immune cells, and endothelial cells. Because of the spatial heterogeneity, bulk sequencing from a specific specimen will produce an average signal of thousands of cells with unknown composition, which forms a hidden confounding factor that interferes with the interpretations of cancer research and diagnosis. Single-cell sequencing inherently has the power to dissect the cellular composition of tissues, providing a powerful tool to advance cancer studies

Fig. 3

Potential applications of single-cell sequencing technologies in cancer research. a Spatial single-cell sequencing. Integration of single-cell sequencing technologies with spatial information of cells to analyze the spatial architecture of tumors. This technique is not yet widely used but is important for cancer biology and treatment. b Single-cell multi-omics. Interrogation of the cellular interaction network within tumors by single-cell sequencing. The very recent development of ProximID, which maps physical cellular interaction networks via single-cell RNA-seq without prior knowledge of component cell types, has proved the principles of single-cell multi-omics [194] and provides great promise for cancer research. c Cellular interaction mapping. Application of single-cell multi-omics techniques to resolve both the somatic mutations and gene expression, which will allow the investigation of immunogenicity of single cancer cells. d Single-cell epigenetics. Techniques to resolve the heterogeneity of cancer cells and tumor-infiltrating immune cells, which may provide new insights into the regulatory mechanisms within tumors and new drug targets to modulate tumor progression