Cancer-associated fibroblast spatial heterogeneity and EMILIN1 expression in the tumor microenvironment modulate TGF-β activity and CD8+ T-cell infiltration in breast cancer

Abstract

Rationale: The tumor microenvironment (TME) and its multifaceted interactions with cancer cells are major targets for cancer treatment. Single-cell technologies have brought major insights into the TME, but the resulting complexity often precludes conclusions on function. Methods: We combined single-cell RNA sequencing and spatial transcriptomic data to explore the relationship between different cancer-associated fibroblast (CAF) populations and immune cell exclusion in breast tumors. The significance of the findings was then evaluated in a cohort of tumors (N=75) from breast cancer patients using immunohistochemistry analysis. Results: Our data show for the first time the degree of spatial organization of different CAF populations in breast cancer. We found that IL-iCAFs, Detox-iCAFs, and IFNγ-iCAFs tended to cluster together, while Wound-myCAFs, TGFβ-myCAFs, and ECM-myCAFs formed another group that overlapped with elevated TGF-β signaling. Differential gene expression analysis of areas with CD8⁺ T-cell infiltration/exclusion within the TGF-β signaling-rich zones identified elastin microfibrillar interface protein 1 (EMILIN1) as a top modulated gene. EMILIN1, a TGF-β inhibitor, was upregulated in IFNγ-iCAFs directly modulating TGFβ immunosuppressive function. Histological analysis of 75 breast cancer samples confirmed that high EMILIN1 expression in the tumor margins was related to high CD8⁺ T-cell infiltration, consistent with our spatial gene expression analysis. High EMILIN1 expression was also associated with better prognosis of patients with breast cancer, underscoring its functional significance for the recruitment of cytotoxic T cells into the tumor area. Conclusion: Our data show that correlating TGF-β signaling to a CAF subpopulation is not enough because proteins with TGF-β-modulating activity originating from other CAF subpopulations can alter its activity. Therefore, therapeutic targeting should remain focused on biological processes rather than on specific CAF subtypes.

Keywords: CAF subpopulations; cancer invasion; patient outcome; spatial transcriptomics; tumor immunity.

Figure 1

BC atlas for cellular and functional annotation of spatial single-cell RNA-seq data. (A) UMAP plot showing the BC atlas based on two previously published single-cell RNA-seq datasets , . (B) Validation of the cellular annotation using several cell-specific genes. (C) Enlarged UMAP plot of CAF subpopulations from the panel (A). (D) (left and center) Scoring of myCAF, iCAF, CAF s1 and CAF s4 signatures in the CAF subpopulations from the panel (C); highest score is denoted in red, lowest score in dark blue. (right) Dot plot of marker genes delineating individual CAF subpopulations; same color-code. Genes used for scoring are outlined in Table S3. (E) Spatial distribution of selected GO processes in BC samples (two representative samples are shown: C1 and 1160920F; other samples are displayed in Figure S13). The following GO processes are displayed: ECM Structural Organization, Wound Healing, TGF-β Receptor Signaling, Regulation of Immune System, Macrophage Activation, and T-Cell Activation.

Figure 2

Spatial analysis of proliferating cancer cells and immune infiltrate in BC samples. (A) Histological annotation of two representative BC samples (other samples are shown in Figure S14), and estimation of highly proliferative regions (S+G2M phases) (higher panels); actively cycling cancer cells and two immune populations (macrophages and CD8⁺ T cells) (middle and lower panels). The heat map shows spatial correlation between these four populations. (B) Correlation analysis for the four selected cell populations in the other eight BC samples.

Figure 3

Spatial relationship between TGF-β signaling and CAF subpopulations in BC. (A) Spatial distribution of genes implicated in TGF-β signaling (top) and spatial distribution of different CAF subpopulations in two BC samples. The dendrogram (bottom, right) shows the spatial co-occurrence between CAF subpopulations and TGF-β signaling. (B) Dendrograms showing the co-occurrence of different CAF populations and TGF-β signaling in the other eight BC samples. (A-B) Labels TGFb, IFNab and IFNg refer to TGFβ, IFNαβ and IFNγ respectively.

Figure 4

Differential gene expression analysis of areas with high TGF-β signaling and with/without CD8⁺ T-cell exclusion. (A) Spatial distribution of areas with high versus low TGF-β signaling and presence/absence of CD8⁺ T cells in two BC samples (patient samples C1 and 1160920F are shown as examples; remaining patients, data not shown). (B) Differential gene expression analysis performed in all 8 BC samples; displayed are the number of patients in which each of the top-modulated genes was found as significantly overexpressed (in the areas where CD8⁺ T cells are present despite high TGF-β signaling). Highlighted in red are EMILIN1 and COL3A1. (C) EMILIN1 expression in the indicated cell subpopulation (from the BC atlas in Figure 1A). (D) EMILIN1 expression in the indicated CAF subpopulations. (E) Upregulation of TGF-β signature genes in the indicated CAF subpopulations. The patient-wise statistical analysis of EMILIN1 overexpression in CD8⁺ cells with high TGF-β signaling regions is provided in Figure S17. (A, D) Labels TGFb, IFNab and IFNg refer to TGFβ, IFNαβ and IFNγ respectively.

Figure 5

EMILIN1 is a good prognostic marker in BC. (A) Multiplexed immunofluorescence analysis displaying the localization of CD8⁺ T cells (red) and EMILIN1 expression (blue) in CAFs in a representative BC sample (N=5). Expression of TGFBI, a TGF-β signaling activity marker, was in green. Cell nuclei were examined by phase-contrast microscopy (data shown in Figure S18). (B) Multiplexed immunohistochemistry analysis showing examples of EMILIN1 (red) and CD8 (brown) co-staining in breast cancer samples (N=75, all subtypes; see also Table S2). (C) Violin plots of CD8⁺ cell counts in areas of high versus low EMILIN1 expression in BC samples (N=75). (D) Ki-67 positivity (as evaluated by retrospective analysis; see Table S4) in EMILIN1-high versus EMILIN1-low cases (N=70); (C-D) p-values were calculated using Mann-Whitney U test. (E) Survival analysis of patients with BC (N=75) in function of EMILIN1 marginal expression level (high versus low). (D-E) The cut-off value of 1 was used to assign patients to EMILIN1 high or EMILIN1 low group. The cut-off value was calculated as ratio of EMILIN1 score in the margin and the score in the central area. Clinical and pathological information regarding the patient cohort are displayed in the Table S2 and S4. Details on scoring methodology are provided in the Materials and Methods section.