Spatial and temporal dynamics of cancer-associated fibroblast niches in breast cancer.

backgroundCancer-associated fibroblasts (CAFs) are the main constituents of the tumor microenvironment. Several studies have delineated CAF heterogeneity in different types of tumors, however, it is still unknown how the distinct CAF transcriptional profiles are established during tumor progression.

methodsWe reanalyzed a previously published single-cell RNA-sequencing dataset of MMTV-PyMT tumors at higher resolution using Seurat , and CytoTRACE to characterize CAF subtypes and their differentiation states. Wilcoxon rank sum test was applied for differential gene expression. Multiplex immunostaining (Akoya PhenoImager HT) was performed on 38 murine mammary tumors from MMTV-PyMT mice to identify the distinct CAF subtypes. Whole-slide imaging and spatial analysis were conducted using QuPath and Cellpose , followed by neighborhood clustering and interaction mapping with CytoMAP . Cellular distances from CAFs to immune, tumor, and endothelial cells were quantified using SPIAT and Wilcoxon tests for comparisons. In parallel, human spatial transcriptomics data from the 10X Genomics Xenium platform were integrated for cross-species validation.

resultsHere, by single cell RNA-sequencing and multiplex immunostaining, we identify six CAF substates. Spatial analysis on immunostained murine mammary tumors and human spatial transcriptomics data outlined temporal changes in stromal composition and the existence of distinct functional niches enriched with different CAF substates. Immunomodulatory CAFs co-localized with immune cells while myofibroblastic CAFs formed a shield around the tumor core, thus preventing immune infiltration.

conclusionsOur work supports the idea that distinct spatial locations dictate different CAF transcriptional programs. Targeting specific functional niches will ultimately hinder tumor progression by inhibiting signaling between distinct CAF substates and the surrounding tumor microenvironment.