The tumor microenvironment (TME) play important roles in progression of endometrial carcinoma (EC). We aimed to assess the cell populations in TME of EC.
We downloaded datasets of single-cell RNA-seq (scRNA-seq) and spatial transcriptome (ST) for EC from GEO, and downloaded RNA-Seq (FPKM) and clinical data of TCGA-UCEC project from TCGA. The datasets were analyzed using R software.
We obtained 5 datasets of scRNA-seq, 1 of ST and 569 samples of RNA-seq. Totally, 0.2 billion transcripts and 33,408 genes were detected in 33,162 cells from scRNA-seq. The cells were classified into 9 clusters, and EC cells were originated from epithelial cells and ciliated cells. Gene set variation analysis (GSVA) indicated that the pathways enriched in the subclusters of epithelial cells and endothelial cells were significantly different, indicating great heterogeneity in EC. Cell-cell communication analyses showed that EC cells emitted the strongest signals, and endothelial cells received more signals than other cells. Further analysis found that subclusters of 1 and 2 of epithelial cells were showed a more malignant phenotype, which may confer malignant phenotype to subcluster of 0 of endothelial cells through MK pathway by MDL-NCL signal. We also analyzed communications between spatial neighbors with ST data and confirmed the findings on MDL-NCL in cell-cell communication. TCGA and GEO analyses indicated that the expression levels of NCL was inversely correlated with ImmuneScore.
Our study revealed EC cells can confer malignant phenotype to endothelial cells by MDK-NCL signal, and NCL is associated with suppressed immune activity. EC cells may shape TME by inhibiting immune cells and “educating” stromal cells