Revelation of comprehensive cell profiling of primary and metastatic tumour ecosystems in oral squamous cell carcinoma by single-cell transcriptomic analysis.

Background: The analysis of single-cell transcriptome profiling of tumour tissue isolates helps to identify heterogeneous tumour cells, neighbouring stromal cells and immune cells. Local metastasis of lymph nodes is the most dominant and influential biological behaviors of oral squamous cell carcinoma (OSCC) in terms of treatment prognosis. Understanding metastasis initiation and progression is important for the discovery of new treatments for OSCC and prediction of clinical responses to immunotherapy. However, the identity of metastasis-initiating cells in human OSCC remains elusive, and whether metastases are hierarchically organized is unknown. Therefore, this study was conducted to understand the cellular origins and gene expression signature of OSCC at the single-cell level. Methods: Single-cell RNA sequencing (scRNA-seq) was used to analyze cells from tissue of para-carcinoma (PCA: adjacent normal tissue not less than 2 cm from the tumour), carcinoma (CA), lymph node metastasis (LNM) from patients with OSCC and PCA and CA tissue from patients with second primary OSCC (SPOSCC) after radiotherapy of nasopharyngeal carcinoma (NPC). The cell types and their underlying functions were classified. The comparisons were then conducted between the homology and heterogeneity from cell types and both conservative and heterogeneous aspects of evolution were identified. Immunohistochemistry was performed to verify the makers of cell clusters and the expression level of novel genes. Results: A single-cell transcriptomic map of OSCC was created, including 16 clusters of PCA cells, 17 clusters of CA cells, 14 clusters of left LNM cells, and 14 clusters of right LNM cells. We also discovered two novel types of cells including CD1C-CD141-dendritic cells and CD1C+_B dendritic cells. Most of the non-cancer cells are immune cells, with two distinct clusters of T lymphocytes, B lymphocytes, CD1C-CD141-dendritic cells+ and CD1C+_B dendritic cells. We also classified cells into 15 clusters for SPOSCC after radiotherapy of NPC. Determining the upregulated expression levels of IL1RN and C15orf48 as novel markers using immunohistochemistry facilitated the correct classification of OSCC including SPOSCC after radiotherapy of NPC and the prediction of their prognosis. Conclusions: The findings provided an unprecedented and valuable view of the functional states and heterogeneity of cell populations in LNM of OSCC and SPOSCC after radiotherapy of NPC at single-cell genomic resolution. Moreover, this transcriptomic map discovered new cell types in mouth, and novel tumour cell-specific markers/oncogene.