Unraveling the Complexity of Prostate Adenocarcinoma: Integrative Insights from Spatial Transcriptome and Single Cell Analysis
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Introduction Prostate adenocarcinoma (PRAD) is a highly prevalent cancer in males, with a dangerously low survival rate for patients with metastatic lesions. The tumor microenvironment (TME), composed of various non-epithelial cell types, plays a significant role in PRAD progression. Incorporating spatial information into the analysis can provide valuable insights for targeted therapies. Methods RNA expression profiles and clinical data were procured from the TCGA-PRAD dataset, scRNA-seq data from the GSE141445 dataset. Spatial transcriptome data, containing two primary tumor samples, was obtained from the 10X Genomics dataset, which were subjected to SpaceRanger software for quality control, normalization, and selection of highly variable genes, and conditional autoregression-based deconvolution (CARD) algorithm for deconvolution analysis. Cell annotation analysis was performed using specific marker genes, followed by subgroup analysis of each cell group using the Sc-Type software. Copy number variation (CNV) analysis and identification of tumor cells were conducted using the Copykat software. Pseudo-time analysis of single cells was performed using the monocle2 software to infer the cell differentiation process. Results We identified PRAD malignant cells and six TME cell types. We identified eight PRAD tumor cell subtypes, labeled C0 to C7. C1 was featured with AMPK and FOXO signaling activation, while C7 was found with cellular senescence and cell cycle, suggesting a malignant cell subpopulation experiencing cell cycle arrest. The benign tissue was mainly composed of C3 and C1, while the proportion of C0 and C2 were significantly elevated in the malignant group. LTB + CD4 T cells and CCL4L2 + NKT cells were highly elevated in the malignant group. ACTA2 was enriched in the early stages of CAF differentiation, while COL1A1 was enriched in the terminal stages. A high rate of overlapping of cancer associated fibroblasts (CAFs) and both cancerous PRAD epithelial and normal cells in 2 PRAD slides was observed, suggesting a possible ongoing carcinogenic role of CAFs. Conclusion We proposed a potential ongoing carcinogenic role of CAFs in PRAD cancer. The comprehensive analysis of both single-cell and spatial transcriptome landscape of PRAD deepens our understanding of PRAD biology and provide valuable insights into the heterogeneity and cellular composition of PRAD tumors. Moreover, we knocked down the expression of CXCL12 in two human prostate cancer cell lines and performed three types of phenotypic experiments. The results validated the role of CXCL12 in prostate cancer, making the bioinformatics conclusions more credible.