Droplet-mediated scRNA-seq reveals cancer and non-cancer components of PPB
To fully understand intra-tumoral and inter-tumor heterogeneity of pleuropulmonary blastoma (PPB), deep transcriptional profiles of the cancer cell status and gene expression were obtained by droplet-mediated scRNA-seq platform (10× Genomics Chromium)14 to observe whether cell diversity existed (Fig.1A). We sequenced 10 007 cells from the individual at an average of 400 M reads. After quality control filtering to remove cells with low gene detection (<500 genes) and high mitochondrial gene coverage (>10%) with the Seurat pipeline, we obtained a rough cell atlas of this sample, which was clustered into 14 clusters (Fig.1D). This whole specimen had several classes of cells, including normal immune cells (cluster 12, PTPRChi), fibroblasts (cluster 13, TAGLNhi ), endothelial cells (cluster 14, CLDN5hi ), and tumor cells (cluster 0-11,NCAM1hiDICER1hiIGF1Rhi )4,15(Supporting Information Figure S1) with differentially expressed genes (DEGs). To further verify such cell classification,TAGLNhi normal fibroblasts from a healthy adult lung were used as a reference to analyze the inferring copy number variation (inferCNV) of all single cells in PPB, because fibroblasts are derived from mesenchyme. We found coherent chromosomal aberrations in tumor clusters (clusters 0–11) but not in the other defined non-cancer cells (clusters 12–14). Tumor clusters contained chromosome 19 gain and chromosome 1, 5, 6, and 19 loss (Fig.1C)15. Interestingly, there were both distinct gain and loss in chromosome 19, which were probably closely connected to tumorigenesis of the chromosome variation.