Tumour-Derived sEVs Promote Triple-Negative Breast Cancer Progression Associated with HAVCR2 Upregulation in Macrophages

Triple-negative breast cancer (TNBC) is the most aggressive breast cancer subtype with a unique tumor microenvironment, and while Programmed cell death protein 1/Programmed cell death ligand 1 (PD-1/PD-L1) blockade represents a standard immunotherapy, most patients develop primary or acquired resistance, with few alternative immunotherapeutic targets currently available. Therefore, we aimed to identify potential immune checkpoint-related molecules involved in TNBC-macrophage crosstalk, clarify the underlying molecular mechanism mediated by small extracellular vesicles (sEVs), and provide a theoretical basis for the future development of novel immunotherapeutic targets against TNBC.

Single-cell RNA-sequencing (scRNA-seq) datasets for various breast cancer subtypes were used. Pseudotime trajectory, cell‒cell communication and Tumor Immune Estimation Resource 2.0 (TIMER2) analyses were conducted to characterize the tumour microenvironment (TME). Immunochemistry and immunofluorescence were used to confirm the results of the above analyses. Single-nucleus RNA sequencing (snRNA-seq) was conducted on three pairs of TNBC tumour and adjacent normal tissues. The functions of tumour-associated macrophages (TAMs) and sEVs in TNBC metastasis were explored by Western blotting, flow cytometry and cell-based experiments.

A total of nearly 60,000 high-quality single cells were subjected to scRNA-seq analysis, from which seven major cell types were identified. An overall increase in immune cell proportion was observed in TNBC compared with other subtypes, with the immune cell fraction in TNBC tissues being ~1.8-fold higher than that in luminal A/HER2+ subtypes (p < 0.001). Cell‒cell communication analysis indicated that TNBC cells mainly interact with macrophages. Interestingly, HAVCR2, an immune checkpoint, is expressed mainly in macrophages in the TNBC TME. HAVCR2 is associated with macrophage pseudotime progression in TNBC, which was validated by immunofluorescence staining. Moreover, analysis of The Cancer Genome Atlas (TCGA) bulk RNA-seq data revealed that HAVCR2 expression is significantly correlated with M2-like macrophage gene signatures and computationally inferred macrophage infiltration levels in TNBC, and this tissue-level transcriptional correlation is associated with poor patient prognosis. Notably, bulk RNA-seq data cannot define discrete cell subsets, and the identification of HAVCR2+ M2 macrophage subsets was independently validated by scRNA-seq and snRNA-seq at the single-cell level. Furthermore, treatment with TNBC-derived sEVs is associated with concurrent increases in the expression of HAVCR2 and M2-associated markers (CD163, CD206) in macrophages. These findings reflect a correlative association rather than a demonstrated causal or regulatory relationship between HAVCR2 and M2-associated marker upregulation.

sEVs derived from TNBC cells are associated with the upregulation of M2-associated markers and concomitant HAVCR2 upregulation in macrophages, both of which correlate with TNBC progression and metastasis. We propose that HAVCR2 may serve as a candidate prognostic marker associated with M2-like macrophage features in TNBC, and these foundational in vitro findings from Human acute monocytic leukemia cell line (THP-1) macrophages warrant further validation in primary human monocyte-derived macrophages and in vivo TNBC models.