Related Experiment Video
Updated: Jan 14, 2026
Optimization of a Multiplex RNA-based Expression Assay Using Breast Cancer Archival Material
Published on: August 1, 2018
Pan-cancer analysis of RNA expression signatures associated with cancer tissue architecture
Megumi Mogi1,2, Katsushige Kawase1,2, Hiroyoshi Y Tanaka3
1Division of Cell Therapy, Chiba Cancer Center Research Institute, Chiba, Japan.
Background:
Tissue architecture of cancer deviates from that of normal tissue and is closely linked to various features of cancer, including invasion and tumor immunity. However, pan-cancer analyses of cancer tissue architecture (CTA) remain limited.
Methods:
We applied non-negative matrix factorization to the expression data of 593 CTA-related genes from 28 cancer types in the Cancer Genome Atlas Project dataset, identifying seven distinct CTA signatures.
Results:
A fibrous collagen-related signature, which is related to fibroblasts and the extracellular matrix, was relatively ubiquitous and represents a universal feature of cancer. In contrast, those associated with cell-cell adhesion or cell-stroma adhesion exhibited high tissue specificity. These findings were validated using Pan-Cancer Analysis of Whole Genomes data, spatial transcriptomics, and in vivo and in vitro models. In renal cell carcinoma, we found a strong correlation between network-forming collagen and patient survival, supported by mouse experiments showing that perturbation of this collagen promotes tumor growth. Additionally, several gene mutations and copy number alterations were found to correlate with changes in CTA.
Conclusion:
Our results suggest that CTA can be understood as a combination of components that influence the pathological features of cancer.
