Fig 1.
Cell clustering and annotation of breast cancer scRNA-seq data.
(A) UMAP plot showing cell cluster annotations across normal breast tissue and breast cancer subtypes. (B) UMAP plot displaying annotations of all identified cell clusters. (C) Dotplot of specific marker genes for each cell subpopulation. (D) GSVA enrichment analysis of all clusters using HALLMARK gene sets. (E) UMAP plot showing subcluster annotation of macrophages. (F) Proportion of macrophage subpopulations across different sample groups. (G) UMAP plot showing subcluster annotation of endothelial cells. (H) Proportion of endothelial cell subpopulations across different sample groups. (I) Heatmap of the R o/e index across cell clusters to quantify tissue preference. (J) Correlation heatmap between cell subpopulations. (K) Cell composition proportions of all identified cell types across grouped samples.
Fig 2.
Inference of cell-type infiltration levels in bulk RNA-seq based on scRNA-seq annotations.
(A) Comparison of infiltration proportions of all annotated cell types between normal and tumor samples in bulk RNA-seq data. (B) Correlation matrix showing the relationships among infiltration proportions of all cell types. (C) Forest plot summarizing Cox regression analysis of the association between cell infiltration proportions and breast cancer survival outcomes. (D) Kaplan–Meier (KM) survival curve for overall survival (OS), comparing high and low CXCL9 ⁺ macrophage infiltration groups divided by the median. (E) Volcano plot showing differentially expressed genes between high and low CXCL9 + macrophage infiltration groups. (F) KEGG pathway enrichment analysis of the differentially expressed genes. (G) Gene Ontology (Biological Process) enrichment analysis of the differentially expressed genes.
Fig 3.
Single-cell transcriptional landscape of macrophage subpopulations.
(A) UMAP plot showing the annotation of macrophage subpopulation. (B) Volcano plot displaying differentially expressed genes for different macrophage subpopulations. (C) Heat map of the R o/e index illustrating tissue preferences of macrophage subtypes. (D) Violin plots comparing the expression levels of CXCL9, SPP1, and MMP9 across macrophage subpopulation. (E) GSVA enrichment analysis of macrophage subpopulations using the HALLMARK gene set. (F) GO (Biological Process) enrichment analysis of highly expressed genes in CXCL9 + macrophages.
Fig 4.
Cell-cell communication analysis among single-cell subpopulations.
(A) Interaction intensity map among all annotated cell types. (B) Heatmap displaying the number of ligand–receptor interactions between cell populations. (C) Interaction intensity map showing communication between CXCL9 + macrophages and other cell types. (D) Dotplot of ligand–receptor interactions between macrophages and T_NK cells within the Th1-related gene family. (E) Chord diagram illustrating the interactions between macrophages and T_NK cells. (F) Dot plot of ligand–receptor interactions between macrophages and epithelial cells within the costimulatory gene family. (G) Chord diagram illustrating the interactions between macrophages and epithelial cells.
Fig 5.
WGCNA analysis and construction of a prognostic risk score based on CXCL9 + macrophage-related genes.
(A) Heatmap showing the correlation between WGCNA modules and cell infiltration traits. (B) Scatter plot illustrating the correlation between module membership and gene significance. (C) GO enrichment analysis of genes in the module most associated with CXCL9 + macrophage infiltration. (D) LASSO regression path plot for selecting survival-associated genes. (E) Ten-fold cross-validation results from LASSO regression showing the optimal lambda value. (F) Kaplan–Meier survival curve comparing high- and low-risk groups based on the median risk score. (G) ROC curves showing the predictive performance of the risk score for 1-year, 3-year, and 5-year survival. (H) Scatter plot showing the relationship between individual risk scores and survival status. (I) Gene interaction network of genes used to construct the risk score model.
Fig 6.
Effects of Macrophage-Conditioned Media on MDA-MB-231 Cells.
(A) Transwell migration and invasion assay of MDA-MB-231 cells co-cultured with conditioned media from M0 macrophages (CON), M1 macrophages (M1), and siCXCL9 M1 macrophages (siCXCL9). (B) Quantification of migrated and invaded cells from the Transwell assay. (C) Wound healing assay showing migration of MDA-MB-231 cells at 0- and 24-hours post-scratch. (D) EdU incorporation assay to assess cell proliferation; nuclei were counterstained with DAPI. (E) Quantification of wound healing, expressed as the ratio of scratch distance at 24 hours to that at 0 hours. (F) Quantification of EdU assay, expressed as the percentage of EdU-positive cells. (G) Colony formation assay. (H) Quantification of colony numbers. Data are presented as mean ± SD. Statistical significance: ns, p ≥ 0.05; * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001.