Abstract P5-04-03: Bone morphogenic protein-4: A novel metastasis suppressor gene in breast cancer

Abstract

Abstract Background: The combination of large-scale screening platforms and animal models of cancer, have provided much insight into the genetic mechanisms that control metastatic progression. Indeed, several studies have unraveled an essential role for genes that encode growth factors and extracellular matrix proteins in the progression of breast cancer. We have found that one such growth factor, bone morphogenic protein-4 (BMP4), to be significantly reduced in tumors with a high proclivity to metastasize. BMP4 is known to regulate tissue polarity and differentiation during embryogenesis, however it is not known whether BMP4 can functionally affect tumor progression. Methods and results: In a panel of mammary tumor lines, we demonstrate an inverse correlation between metastatic propensity and the expression of BMP4 through a combination of RT-PCR, immunohistochemistry (IHC) and ELISA. These findings were extended to publicly available gene expression data sets, where low BMP4 expression was found to be associated with ER-negative breast tumors and in those tumors with high histologic grade. Low BMP4 expression also correlated with poorer survival from distant metastases (HR 0.82, p = 0.013). IHC analysis on a tissue microarray consisting of tumor specimens from 535 patients with invasive breast cancer demonstrated that, compared to normal breast epithelium, BMP4 positivity was significantly less common in both DCIS (HR 0.59, p = 0.00046) and invasive carcinoma (HR = 0.56, p<0.0000001), and was inversely associated with axillary lymph node-positivity (HR = 1.53, p = 0.055). Using surrogate in vitro assays of metastasis, we determined that BMP4 can suppress the ability of highly metastatic 4T1.2 tumor cells to resist anoikis. When BMP4 was overexpressed in 4T1.2 cells (4T1.2-BMP4) and orthotopically implanted in mice, we did not observe an effect on primary tumor growth, however elevated BMP4 expression did block the ability of these tumors to metastasize to the lymph node, lung and bone. In a reverse-complimentary approach, we confirmed that silencing of BMP4 expression by RNAi in weakly metastatic 4T07 and 168FARN cells, can enhance lung colonization. Mechanistically, we establish that BMP4 can induce canonical BMP-SMAD signaling in multiple breast cancer cells, leading to an up-regulation of genes known to suppress metastasis, and a down-regulation of metastasis promoting genes. Specifically, we link the anti-metastatic function of BMP4 to its ability to induce the expression of the known metastasis suppressor, Smad7. Through RNAi-mediated suppression of Smad7 in 4T1.2-BMP4 tumors, we were able to restore the metastatic phenotype of this tumor line. Finally, we demonstrate that administration of recombinant BMP4 protein in 4T1.2 tumor challenged mice elevates Smad7 expression within the primary tumor, and leads to a pronounced decrease in spontaneous bone and lung metastasis. Conclusion: Utilizing animal models of cancer, and clinically sourced tissues, we provide correlative and functional data to prove that BMP4 is a bona fide metastasis suppressor gene in breast cancer. Furthermore, we demonstrate that BMP4 may be therapeutically viable, and can prevent breast tumor progression through the modulation of known ‘metastasis virulence’ genes. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P5-04-03.