Integrated analyses of single-cell and bulk transcriptome reveal cell type-specific metastasis- and prognosis-related genes in breast cancer

Abstract

Abstract Background Single-cell technologies raise the possibility of providing type-specific insights into tumor microenvironment and facilitate the development of personalized medicine. The object of this research was to afford a novel means to explore the cell type-specific differentially expressed genes (DEGs) between primary cancer and metastatic lymph nodes which were also utilized to investigate the potentials of clinical practice. Methods We collected single-cell and bulk transcriptome sequencing data from two public databases. With single-cell data, we explored the cell type-specific DEGs between primary cancer and metastatic lymph nodes. Also, the cell type-specific DEGs among different states through single-cell pseudotime analysis were identified as the potential genes cardinal for cell differentiation. With the intersection DEGs and bulk transcriptome sequencing data, we further delineated the therapeutic potentials of these DEGs through identification of a prognostic signature which could be used to facilitate the stratification of patients with different outcomes. We also investigated the different cell communication patterns between primary cancer and metastatic lymph nodes. Results We identified 2177 cell type-specific DEGs between primary cancer and metastatic lymph nodes. We further identified 2330 cell type-specific DEGs among different states through single-cell pseudotime analysis. The intersection DEGs were incorporated into bulk transcriptome sequencing data, with which we constructed a signature comprising of eight genes and validated it using an independent cohort. The samples with high-risk also exhibited low levels of immune infiltration compared to high-risk samples. The cell interactions in metastatic lymph nodes were mainly downregulated except macrophage migration inhibitory factor (MIF) signal pathway. Conclusion The cell type-specific DEGs identified though single-cell data might be the potential therapeutic targets. The robust signature could be used to predict outcomes of patients especially in combination with conventional TNM stages. We also demonstrated the benefits of immune infiltration in breast cancer. The exclusive MIF signal pathway in metastatic lymph nodes might be correlated with the metastasis and deserved more studies.