A spatial transcriptome-based perspective on highly variable genes associated with the tumor microenvironment in hepatocellular carcinoma
Hang Yang, QiNian Jiang
Abstract
Hepatocellular carcinoma is a common pathologic and histologic subtype ofprimary liver cancer and a common cause of cancer-related death. The tumor microenvironment (TME) is involved in the composition and deterioration of the cancer pathological environment and thus understanding the components of the TME of hepatocellular carcinoma and the mechanisms involved will aid in the developmentof strategies for inhibiting hepatocellular carcinoma progression and metastasis. The spatial transcriptomic data were used to uncover specific cell subclusters based on Seurat’s process, followed using Monocle to perform pseudotemporal analysis. Spatial gene pathways were analyzed using the gene set variation analysis(GSVA) package process, and single-cell sequencing data were processed using Seurat. Gene sets were scored using the AddModuleScore approach. Quantitative polymerase chain reaction and immunohistochemistry as well as molecular docking assays were used to determine the biological effects of S100A6. The tumor tissueregion revealed a specific subpopulation called cluster 4, which was rich in cancer associated fibroblasts, particularly myofibroblasts. GSVA analysis suggested that this subpopulation was involved in the epithelial-mesenchymal transition process, and Kyoto Encyclopedia of Genes and Genomes analysis uncovered its involvement in the composition of extracellular matrix components. Pathological sections of the spatial transcriptome revealed that this subpopulation was specifically expressed at the tumor’s broad envelope boundary and in stroma-rich locations, as confirmed with subsequent sections. The most prognostically relevant gene, S100A6, wasdetected using TCGA sequencing, and expression patterns and tissue locations were consistent with the similar trends of this subpopulation. In conclusion, we identified subgroups with tissue-specific expression patterns and marker genes that are most substantially associated with cancer prognosis, allowing us to gain a more comprehensive understanding of the intricate compositional alterations in the TME.