BackgroundHead and neck squamous cell carcinoma (HNSCC) accounts for more than half of head and neck tumors, roughly 90%. This study focused on constructing the heterogeneous landscape using single-cell and bulk transcriptomic data to identify molecular characteristics of the microenvironment in primary and lymphatic metastatic head and neck squamous carcinomas. MethodThe study enrolled 23 HNSCC samples with scRNA-seq data and 546 HNSCC samples from TCGA. We used Monocle to sort the cells and used CellPhoneDB to explore the cell-cell interactions. Infercnv, which was used to infer cells with apparent copy number variation based on single-cell sequencing transcriptome data. We re-evaluated HNSCC bulk RNA transcriptome data to characterize the functions of different cell types in shaping the immune microenvironment of HNSCC. ResultsWe combined genealogical reconstruction, CNV inference, and cellular interactions to uncover the characteristics of distinct cell populations in different disease states, differences in cancer and immune cell lineages of differentiation trajectories, and interactions between non-immune and immune cell. PD-1 and PD-L1/PD-L2 expressed extremely rare in T cells, the immune checkpoint molecule KLRB1-CLEC2D achieved a high-level expression. We identified three microenvironment-based HNSCC subtypes associated with the prognosis of HNSCC patients. ConclusionsIn summary, the present study dissected the intratumoral heterogeneity and immune microenvironment of primary and metastatic HNSCC, which is crucial to reveal the mechanisms of resistance to immunotherapy in HNSCC in different disease states and is expected to assist in the further investigation of the mechanism of HNSCC cell metastasis and guide the treatment of clinical patients.