The Expression Signatures of Mammary Tumor Associated Macrophages.

It is well established that an increased density of tumor-associated macrophages (TAMs) correlates with poor prognosis in many types of solid tumors. This evidence is particularly strong for breast cancers. A causal relationship between TAMs and poor prognosis was suggested by experiments whereby a genetic depletion of macrophages in a mouse model of breast cancer caused by the mammary restricted expression of the Polyoma Middle T oncoprotein (PyMT) slowed tumor progression and inhibited metastasis. Subsequent studies in these primary mammary tumors showed that TAMs directly or indirectly promote tumor angiogenesis as well as tumor cell migration, invasion and intravasion. TAMs are also thought to affect the inflammatory context of the tumor microenvironment by suppressing adaptive immune responses that would normally reject the growing tumor. The varied tasks ascribed to TAMs suggested that the tumor microenvironment educates different population of macrophages to perform specific tasks. In this study, we isolated TAMs from the PyMT primary tumors in order to evaluate their gene expression signatures compared to a resident splenic population in order to define specific tumor associated functions. To perform these studies mice that express enhanced green fluorescent protein (eGFP) from the colony stimulating factor 1 receptor (CSF-1R) mononuclear phagocytic restricted promoter were crossed to the PyMT animals to generate offspring with eGFP+ TAMs. Animals were injected with dye-conjugated dextran two hours prior to sacrifice in order to identify phagocytic cells, a characteristic of macrophages. We have established that these eGFP+/dextran+ cells are F4/80+ and define the TAM population. EGFP+/dextran+ TAMs were isolated using flow cytometry from late-stage tumors and splenic macrophages were sorted from non-tumor bearing animals using an identical protocol. These two populations were analyzed on gene expression oligoarrays to better elucidate specific mediators of TAM pathogenicity. We have identified approximately 100 genes whose transcript abundance are up or down regulated in the TAM population including genes mediating angiogenesis, adhesion and inflammation. Furthermore, genes previously described to define the tumor associated suppressor macrophage (MIF-1, MIP1α and TGFβ, high; IL-18, low) were similarly regulated amongst the three biological repeats. To further define individual TAM populations, we used an in vivo invasion assay to isolate a subset of TAMs that promote carcinoma cell motility in vivo. This assay involves the collection of invasive tumor cells and co-migrating invasive TAMs into EGF-containing microneedles placed directly into the primary tumor of an MMTV-PyMT animal. Previously, this assay was used to describe a paracrine loop in which carcinoma cells secrete CSF-1 that binds CSF-1R on TAMs leading to TAM secretion of epidermal growth factor (EGF) that binds the EGF receptor on carcinoma cells and stimulate their motility. Disrupting this paracrine loop is known to block the invasion of both cell types. Invasive TAMs isolated via this assay and separated from invasive carcinoma cells using CD11b magnetic beads were compared by gene expression arrays to TAMs sorted by flow cytometry (F4/80+/dextran+). The transcript abundance of about 200 genes were differentially regulated between these two populations. Together, these two studies illustrate key genes expressed in TAMs that may regulate tumor progression and furthermore, define a specific sub-population of TAMs that directly promotes tumor cell migration and invasion.