Breast Cancer Stem Cell Self-Promotion Occurs through Extracellular Matrix Modification

Since the “War on Cancer” was declared in 1971 by the signing of the National Cancer Act, many advances have improved our understanding of this complex and heterogeneous collection of diseases. Standard of care treatments generally focus on reducing bulk tumor cells by eliminating rapidly dividing cells. Nevertheless, tumors often recur even after long periods of remission, suggesting that small populations of tumor cells called cancer stem cells (CSCs) are resistant to conventional anti-tumor therapies and retain the capacity to self-renew and differentiate. Newer lines of research are focused on understanding CSC biology so that therapies targeting them can be developed. In this month’s issue of Genes and Development, a cross-institutional group led by Drs. Hira Lal Goel and Arthur Mercurio at the University of Massachusetts Medical School offer data demonstrating that laminin 551, an extracellular matrix (ECM) protein, is produced by breast CSCs and is required for these cells to retain their self-renewal and tumor initiation capabilities (Chang, et al. 2015).

Alternative splice isoforms of extracellular matrix proteins determine cancer stem cell properties

In a previous study, the Goel/Mercurio group identified two sub-populations of CD44high/CD24low cells derived from SRC-transformed human breast cancer cells that differed in both α6β1 integrin splice variant expression and morphological characteristics (Goel, et al., 2014). The population expressing α6Aβ1( α6A) were epithelial in appearance and more phenotypically stable, whereas the α6Bβ1( α6B)-expressing population were mesenchymal, and displayed more CSC-like properties – specifically, they proliferated more aggressively, and were able to self-renew. Importantly, only the α6B-positive population induced tumors when injected into the mammary fat pads of immunodeficient NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ (NSG) mice (005557), indicating that α6B could be a functional marker for breast cancer stem cells in vivo.

Interestingly, these α6A and α6B integrin variants, which serve as laminin receptors, differ in their cytoplasmic domains, suggesting that each splice variant might activate different signaling pathways upon binding to a laminin matrix. The discovery of a specific differentially spliced α6 integrin variant in putative CSCs suggested also that a particular laminin in the tumor ECM might interact preferentially with CSCs to sustain their stem cell properties. Indeed, in their most recent publication, the Goel/Mercurio group found that laminin 511 is the preferred ligand for α6B (Chang, et al., 2015), supporting this hypothesis that the extracellular matrix milieu might affect cancer stem cell properties.


NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ (005557)

Expression of extracellular matrix proteins is driven through a positive feedback loop

While examining the phenotypic differences between α6A and α6B breast cancer cells using RNA sequencing (RNA-Seq) analysis, the Goel/Mercurio group found that laminin α5 (LMα5), a subunit of laminin 511, is preferentially expressed in α6B cells. In both their RNA-Seq and subsequent immunofluorescence analysis, they also found that the Hippo transducer, TAZ, which has been reported to promote cancer stem cell properties in breast cancer (Cordenonsi et al. 2011), was expressed at higher levels in the nucleus of α6B-expressing cells, and induces LMα5 transcription. Together these observations suggest that the “stemness” of α6B-expressing CSCs might be sustained through self-reinforcing interactions with their extracellular matrix via a TAZ-mediated positive feedback loop.

Finally, the Chang et al. studies demonstrated that α6B and LMα5 are clinically relevant and function in mouse models. From triple negative patient breast tumors, the investigators identified a small population of LMα5-positive cells that exhibited enhanced stem and progenitor cell-like characteristics when compared to LMα5-negative populations. Importantly, NSG mice engrafted with α6B-expressing cells with enforced reduction of LMα5 exhibited longer tumor latencies than controls engrafted with unmodified α6B-expressing cells. The authors also confirmed the contribution of α6B and LMα5 to tumor formation in a mouse transgenic model of spontaneous triple negative breast cancer that they had generated previously (Kumar et al., 2012, Goel et al., 2013). Like the cells derived from the patient tumors, α6B-expressing CSCs isolated from the tumors of these “TBP” mice (TgMFT121; Brca1flox/flox p53flox/flox; TgWAP-Cre), showed a longer latency to form tumors in vivo when their LMα5 levels were artificially reduced.


Together, the results reported in Chang et al. define intriguing, self-reinforcing interactions between CSCs and their underlying ECM that are required to sustain their proliferative and tumor-initiating potential. Further, this study adds credibility to theory that CSCs contribute to tumor formation and disease relapse in cancer, and points to the tumor microenvironment - and the extracellular matrix in particular - as a potential target for new therapeutic treatments.