Abstract Background: A subpopulation of tumor cells known to have intrinsic resistance to therapies and contribute to metastasis function as “cancer stem”, or tumor-initiating, cells (TICs). TICs can be identified by various biomarkers, including fluorescent reporters for signaling pathways that regulate TIC function such as STAT signaling. However, standard TIC reporters have limitations. Their expression is unstable and there is no established method to follow TICs as they undergo cell state changes. Furthermore, we do not completely understand the transcriptional features of TICs. Improved methods of identifying TICs as well as the development of high-resolution transcriptional signatures are essential to understand TIC biology. An improved understanding of TIC biology can inform the development of novel therapeutics to selectively ablate the TIC compartment, thus improving response to standard cancer therapies. Methods: We screened a panel of PDX models to determine whether a STAT reporter identifies TICs in patient-derived xenograft (PDX) models. To augment existing STAT TIC reporters, we developed a two-component STAT signaling-specific lineage-tracing (LT) system. The first component labels active STAT signaling cells by expression of enhanced green fluorescent protein (EGFP+), followed by a self-cleaving P2A peptide and a TAM-inducible Cre-recombinase (4M67-EGFP-P2A-CreERT2). The second component is a constitutively expressed dual-color switching Cre-dependent reporter vector (EFS-loxPdsRedloxP-mNeptune2). We determined which LT cell populations (EGFP+/mNeptune2+, EGFP+/dsRed+, EGFP−/mNeptune2+, and EGFP−/dsRed+) from SUM159 xenografts functioned as TICs. To probe TIC transcriptional features, we performed scRNA seq on the two tumor models with STAT TICs and a PDX model that did not have STAT TICs. We performed lineage trajectory analysis and differential gene expression analysis to characterize candidate TIC transcriptional signatures. To confirm whether these transcriptional features are associated with TICs, we identified a biomarker of this cell state (bone marrow stromal antigen 2 [BST2]) and explored the relationship between this biomarker and TICs. Results: We identified four PDX models with STAT reporter activity, one of which had STAT TICs. We validated our LT system in several models both in vitro and in vivo, then demonstrated EGFP+/mNeptune2+ from SUM159 xenografts are enriched for TICs and that lineage-tagged cells (mNeptune2+) function as proliferative early progenitor cells. scRNA seq of three xenograft models uncovered a distinct antiviral cell state in all three models that represent a more highly enriched TIC subpopulation. Critically, transcriptional features were shared between the antiviral cell states across xenograft models. To determine whether cells in this antiviral state represent TICs, we FACS-enriched BST2 cells (an IFN-stimulated cell surface protein that was a marker of the antiviral cell state) and demonstrated enrichment of the TIC population by mammosphere formation assays and limiting dilution transplantation. We demonstrated that BST2 expression is increased in residual PDX tumors following chemotherapy. Furthermore, we demonstrated similar antiviral cell states are present in human breast cancer scRNA seq datasets. Conclusions: These data demonstrate TICs adopt an antiviral cell state in some triple-negative breast cancers, and BST2 is a function marker of TICs in this cell state. Therefore, targeting genes associated with this cell state or associated antiviral pathways may represent a therapeutic vulnerability to eliminate TICs in some breast cancers. Citation Format: Eric Souto, Ping Gong, John Landua, Ram Srinivasan, Lacey Dobrolecki, Abhinaya Ganesan, Michael Lewis. Interferon-induced bone marrow stromal antigen 2 (BST2) is a functional tumor-initiating cell marker in triple-negative breast cancer [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO5-24-02.
Read full abstract