Transcription factor ATF3 links host adaptive response to breast cancer metastasis

Host response to cancer signals has emerged as a key factor in cancer development; however, the underlying molecular mechanism is not well understood. In this report, we demonstrate that activating transcription factor 3 (ATF3), a hub of the cellular adaptive response network, plays an important role in host cells to enhance breast cancer metastasis. Immunohistochemical analysis of patient tumor samples revealed that expression of ATF3 in stromal mononuclear cells, but not cancer epithelial cells, is correlated with worse clinical outcomes and is an independent predictor for breast cancer death. This finding was corroborated by data from mouse models showing less efficient breast cancer metastasis in Atf3-deficient mice than in WT mice. Further, mice with myeloid cell-selective KO of Atf3 showed fewer lung metastases, indicating that host ATF3 facilitates metastasis, at least in part, by its function in macrophage/myeloid cells. Gene profiling analyses of macrophages from mouse tumors identified an ATF3-regulated gene signature that could distinguish human tumor stroma from distant stroma and could predict clinical outcomes, lending credence to our mouse models. In conclusion, we identified ATF3 as a regulator in myeloid cells that enhances breast cancer metastasis and has predictive value for clinical outcomes.

Abstract 2843: Tumor macrophages utilize ATF3 to promote breast cancer metastasis

Objective: Cancer progression is facilitated by a complex network of interactions between cancer cells and host-derived components, such as stromal cells and extracellular matrix in the tumor microenvironment. We sought to identify the factors that mediate these dynamic cancer-host interplays. Specifically, we asked how the host responds to signals from the cancer cells. To this end, we investigated the role of ATF3, an adaptive-response gene in the cellular stress response network. Overwhelming evidence indicates that ATF3 is induced by a broad spectrum of extra- and intra-cellular signals in a variety of cell types. As such, it is as an excellent candidate for mediating host responses to cancer cells.

Methods: We injected breast cancer cells (MMTV-PyMT cells) into syngeneic wild type (WT) or ATF3 knockout (KO) mice, performed survival surgery to remove the primary tumors, and examined metastasis two months after tumor removal.

Results: We found that ATF3 deficiency in the host did not affect primary tumor formation; excitingly, though, it dramatically decreased lung metastasis. Analyses of the circulating tumor cells and lung colonization indicated that both early and late steps in the metastatic cascade were defective in the KO host. Since soluble factors are an integral part of the mechanisms by which the host transmits systemic responses, we analyzed the plasma of normal and tumor-bearing mice by an antibody array. Intriguingly, plasma from tumor-bearing KO mice demonstrated a marked impairment in the abundance of various molecules that are known to play important roles in metastasis, suggesting that ATF3 in the host promotes a systemic environment that enhances cancer metastasis. Since the KO mice are whole body KO, the results above do not indicate the cell type(s) in which ATF3 is playing this critical role. We will present evidence that ATF3 plays an important role, at least in part, in the tumor associated macrophages (TAMs), where it up-regulates MMP9 as a functionally important target gene. Two lines of evidence indicated that our findings on ATF3 have clinical relevance. First, analyses of human tumor microarrays by immunohistochemistry revealed that ATF3 expression in monocytic cells correlated with poor outcome. Second, analyses of the mouse TAMs from the WT and KO host identified ∼400 ATF3-regulated genes. Among these genes, a 60-gene signature was identified that could distinguish the human breast tumor stroma from the normal breast stroma (McGill Breast Stroma dataset). Significantly, this signature predicted outcome in two independent patient cohorts.

Conclusion and Significance: We uncovered a previously unknown role for ATF3: it is induced in the host during cancer development and its expression in the host cells, specifically the TAMs, promotes metastasis. This finding is significant because it not only links host stress response to cancer metastasis, it also identified a new gene signature that predicts outcome.

Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 2843. doi:10.1158/1538-7445.AM2011-2843

Vascular endothelial growth factor secretion by tumor-infiltrating macrophages essentially supports tumor angiogenesis, and IgG immune complexes potentiate the process

Tumor growth requires neoangiogenesis. Members of the vascular endothelial growth factor (VEGF) family play an important role as angiogenic promoters in malignant tumors. Tumor cells and stromal cells are sources of VEGF in the tumor. We tested the relevance of the tumor-infiltrating macrophage (TIM) contribution as a source of VEGF in the tumor environment and the role of the local immune complexes in inducing the TIM release of VEGF. Colon and breast carcinoma biopsies were studied with immunoperoxidase staining of CD11b, sialyl-Tn (sTn) antigen (Ag), and gamma immunoglobulin (IgG). The presence of TIM containing phagosomes positive for both IgG and sTn Ag was observed in all tumors, showing that TIMs endocytosed local immune complexes. Reverse transcription-PCR analysis of macrophage (MO) mRNA showed VEGF-A and -B, but not VEGF-C or -D. That pattern was not modified by the presence of tumor cells. In vitro, the interaction of tumor cells and MO promoted the secretion of MO VEGF. The MO secretion of VEGF was augmented when tumor cells were added to cocultures containing MOs and polymorphonuclear cells. Immune complexes formed with tumor sTn Ag and IgG induced a 5-fold increase of MO VEGF secretion. In vivo, TIMs and neoangiogenesis were associated. In vivo experiments with severe combined immunodeficient and athymic nude (nu/nu) mice showed increased number of TIMs, increased tumor angiogenesis, and faster tumor growth in mice with significant serum anti-sTn IgG. This study demonstrates that VEGF secreted by TIMs represents an essential support for tumor angiogenesis and growth, certainly influenced by the humoral antitumor immune response.

IgG-recognizing shed tumor-associated antigens can promote tumor invasion and metastasis

Tumors secreting glycoproteins that act as tumor-associated antigens have been described as highly invasive and metastatic. In this study, the consequences of the humoral immune response (HIR) against these antigens were investigated. Using an in vitro model of tumor cell invasion, results indicated that the invasiveness of tumor cells secreting antigenic secreted/shed tumor glycoproteins (STGP) increases in the presence of specific anti-STGP IgG, polymorphonuclear cells and monocytes. This in vitro model showed that the coincidental presence in the matrix of both STGP and specific anti-STGP IgG increases the local release of IL-1beta, IL-6 and vascular endothelial growth factor (VEGF) by stromal cells, but not by tumor cells. Using an in vivo model, the experiments show that immune-competent mice develop an anti-tumor HIR with anti-STGP IgG production. In this model, tumor growth was increased in parallel with the serum concentration of specific anti-STGP IgG. In athymic nude (nu/nu)-beige mice the same trend was observed, suggesting a T-cell-independent tumor-promoting effect induced by anti-STGP IgG. Tumor histology showed intense infiltration of IgG-positive plasma cells and lymphocytes. A severe combined immunodeficient-beige mouse-based in vivo model of tumors, experimentally infiltrated with monoclonal IgG plasmocytoma cells, showed that only specific anti-STGP-IgG-secreting cells could exacerbate tumor invasion, angiogenesis and metastasis. These results suggest that tumors shedding/secreting antigenic STGP can induce a host IgG immune response that can promote invasion and metastasis by inducing tumor infiltrating stromal cells to release proinflammatory cytokines and VEGF.