Abstract P6-12-14: Histone deacetylase inhibitor suberoylanilide hydroxamic acid targets breast cancer stem cells and inhibits metastasis

Introduction: Inflammatory breast cancer (IBC) is a distinct and aggressive subtype of locally advance breast cancer associated with increased aldehyde dehydrogenase 1 (ALDH1) positive cancer stem cells (CSCs). IBC is associated with a poor survival rate (40% 5-year survival), with few therapeutic strategies identified that effectively blocked the growth and metastasis of IBC. Our previous in vitro studies revealed that the pan-histone deacetylase inhibitor Suberoylanilide Hydroxamic Acid (SAHA) effectively targets self-renewal of IBC tumor cells.

Materials and Methods: The IBC cell line, SUM149 was tagged with the Luciferase gene (SUM149-LUC) and their in vivo growth in immune compromised mice tested in both orthotopic and metastatic settings.

Results: SUM149-LUC rapidly develop primary tumors and metastatic lesions at a variety of locations including lung, brain, bone, liver, reproductive organs and adrenal glands. SAHA effectively blocked growth of SUM149 primary tumors and inhibited metastasis, which was synergistic with the microtubule stabilizer, Paclitaxel. The therapeutic efficacy of SAHA was associated with a significant decrease in ALDH1+ CSCs populations.

Conclusions: Collectively, these results suggest that strategies of combining agents such as SAHA that target CSCs with Paclitaxel that targets the bulk of proliferating tumor cells warrant further investigation for their potential effectiveness in IBC patients, who have the lowest survival of breast cancer patients and have few therapeutic options.

Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P6-12-14.

Abstract P5-04-05: E-Cadherin is Required for In Vivo Growth of Inflammatory Breast Cancer: Importance of Mesenchymal-Epithelial Transition (MET) and Role of HIF1α

Background: Inflammatory breast cancer (IBC) is the most aggressive and deadly form of breast cancer and is phenotypically distinct from other forms of locally advanced breast cancer. The 5-year survival for IBC patients (40%) has not improved even with the implementation of multi-disciplinary care and targeted therapies (e.g. anti-HER2 or anti-estrogen therapies). Lymph node involvement is common at first diagnosis and the accelerated metastasis program that characterizes IBC account for a disproportionate number of the 40,000 women who will die from breast cancer. The most common molecular marker for IBC is the cell adhesion surface glycoprotein, E-Cadherin. Classically, E-Cadherin is thought to be a tumor suppressor and its loss is associated with the epithelial-mesenchymal transition (EMT) and acquisition of an invasive phenotype. The importance of EMT in metastasis is well accepted in numerous tumor models although the presence of EMT in patient samples is rarely demonstrated. Given the propensity of IBC to metastasize, one would expect a strong EMT signature in IBC cell lines, however the persistent robust expression of E-Cadherin in IBC patient tumors and in IBC cell lines, with a lack of expression of ZEB1, a transcriptional repressor of E-cadherin and diminished expression of genes within the transforming growth factor beta signaling pathway, suggests otherwise.

Methods: To further define the role of E-Cadherin in IBC, we used the SUM149 IBC cell line to manipulate E-Cadherin via shRNA knockdown and by overexpression of ZEB1 a known transcriptional repressor of E-cadherin.

Results: While the lack of E-Cadherin or ZEB1 overexpression in SUM149 cells led to an EMT phenotype in vitro, E-Cadherin was required for in vivo growth of SUM149 tumor cells. Gene profiling identified novel E-Cadherin signaling pathways including an hypoxia gene signature through HIF1a as well as cytokine responsive pathways. The requirement of HIF-1a for in vivo growth of SUM149 cells was confirmed by shRNA knockdown of HIF-1a.

Conclusions: The absolute requirement for E-Cadherin and maintenance of the mesenchymal-epithelial transition in IBC tumor growth is demonstrated. Furthermore, a novel functional link between E-Cadherin and HIF-1a and in their critical role in survival of IBC tumors is identified for the first time and warrants further investigation.

Supported in part by The Susan G Komen Organization Promise Grant KG081287 (FMR and MC)

Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P5-04-05.

Abstract P4-06-03: Zinc Finger Nuclease Genome Engineering Reveal Multiple Functions of Secretory Leukocyte Peptidase Inhibitor in Regulating Pleuripotency of Cancer Stem Cells in Inflammatory Breast Cancer

Background: Inflammatory breast cancer (IBC) is the most aggressive and lethal variant of this disease and is known to be enriched for cells with a cancer stem cell phenotype. IBC is characterized by the presence of cell aggregates, defined as tumor emboli, that metastasize into skin and chest wall. The only documented biomarker of tumor emboli is the surface glycoprotein, E-cadherin. We previously demonstrated that IBC tumor emboli express the alarm anti-protease, secretory leukocyte peptidase inhibitor (SLPI), a metastasis related gene highly expressed in IBC patient tumors. Since the function of SLPI in IBC is unknown, the present studies used zinc finger technology to knockout (KO) copies of SLPI in the SUM149 IBC cell line to define the role of SLPI in IBC.

Materials and Methods: Using CompoZr zinc finger nuclease (ZFN) technology (Sigma-Aldrich), SLPI KO cell lines were generated by disrupting all alleles (3) in exon 1 of SUM149 cells derived from an IBC primary tumor with a high CD44+cell population. The target-specific ZFNs bound DNA at a sequence-specific location and created double strand breaks repaired by non-homologous end joining, resulting in deletions at the SLPI locus. Single cell SLPI KO clones were isolated and serially passaged to establish stable cell lines. Functional assays were used to assess proliferation, invasion, tube formation and clonogenicity. Global transcriptional profiling was performed to identify genes and signaling pathways directly regulated by SLPI.

Results: SUM149 SLPI KO clones did not produce SLPI protein and had a significantly slower turn-over time of 75 hrs compared with 24 hrs in SUM149 wild type clones. Loss of SLPI blocked invasion by 50%, and completely inhibited formation of tube-like structures, an activity defined as vasculogenic mimicry, characteristic of IBC. The loss of only 1 SLPI allele resulted in the inability of SUM149 cells to grow as anchorage independent clones in soft agar, commonly used as a predictor of in vivo tumorigenicity. SLPI directly regulated expression of multiple genes within the embryonic stem cell pleuripotency canonical pathway, including WNT, Frizzled, PDK-1, platelet derived growth factor receptor and sphingosine-1-phosphate receptor. Studies are underway to determine the specific role of SLPI in IBC tumor growth and metastasis.

Conclusions: Our previous studies demonstrated that SLPI is expressed by IBC tumor emboli and can be used as a biomarker of tumor emboli in IBC core and skin punch biopsies. SLPI was found to not only regulate critical functional activities of IBC tumor cells but also to directly regulate genes within pathways critically important to maintenance of pleuripotency of tumors with a cancer stem cell phenotype. Collectively, these studies demonstrate the power and utility of the zinc finger technology, which enables the interrogation of tumor cells to discern the direct function and role of specific genes of interest.

Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P4-06-03.

P2-05-04: Mapping the Specific Gene Families Activated in the Lymphangiogenesis and Vasculogenic Mimicry Exhibited by Inflammatory Breast Cancer

Background: Inflammatory breast cancer (IBC) is the most metastatic variant of locally advanced breast cancer. Although IBC is diagnosed less commonly than other types of breast cancer, it is extremely aggressive, and accounts for a disproportionate number of breast cancer related deaths annually. IBC exhibits very specific patterns of lymphangiogenesis and vasculogenic mimicry, however detailed studies of the genes and proteins involved in these angiogenic processes are lacking. This study performed whole unbiased gene transcription studies with validation by protein arrays using all available pre-clinical cell lines and in vivo xenograft models of IBC, including a new model of IBC, FC-IBC01, which exhibits lymphovascular invasion, to identify the specific pathways involved in the distinctive angiogenesis observed in IBC.

Materials and Methods: Real-time quantitative RT-PCR, cDNA microarray gene profiling, immunofluorescence with confocal imaging and protein arrays were used to examine differential expression of specific angiogenic gene families including VEGFA,B,C,D, VEGF Receptor genes, and ANG/TIE genes linked to angiogenesis and lymphangiogenesis.

Results: Activity of the matrix metalloproteinase, MMP-2, is required for IBC tumor cells to undergo vasculogenic mimicry (VM), which is associated with a loss of TIMP-2, a well known inhibitor of angiogenesis. Therapeutics that target MMP activity can successfully inhibit this VM. Furthermore, pre-clinical models of IBC that form IBC tumor emboli exhibit lymphovascular invasion that is associated with distinct patterns of expression of genes that encode for distinct receptor tyrosine kinases that may represent important therapeutic targets for IBC.

Discussion: Identification of the distinct angiogenic pathways that are activated in IBC provides insight into the therapeutic targets that may abrogate the distinct lymphovascular invasion and vasculogenic mimicry that are linked to the aggressive metastasis of IBC.

Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P2-05-04.

P1-02-03: The Reciprocal Roles of E-Cadherin and ZEB1 Demonstrate the Mesenchymal-Epithelial Transition as a Primary Characteristic of Inflammatory Breast Cancer

Background: Inflammatory breast cancer (IBC) is a rare but very aggressive form of breast cancer. IBC is characterized by nests of tightly aggregated cells, defined as tumor emboli, that exhibit characteristics of cancer stem cells (CSCs). IBC tumor emboli express E-cadherin which is required to maintain their integrity and our recent evidence demonstrates that expression of E-cadherin by tumor emboli is associated with lack of ZEB1 expression, a transcriptional repressor of E-cadherin. This is at odds with the current hypothesis that metastatic progression is associated with the process of epithelial mesenchymal transition (EMT), with loss of E-cadherin and gain of transcription factors including ZEB1, acquisition of CSC characteristics and enhanced invasive capabilities.

Materials and Methods: shRNA knockdown and over-expression methods, real time PCR arrays, western blotting, and in vitro assays to evaluate proliferation, invasion, growth in soft agar and clonogenicity and in vivo animal studies were used.

Results: Expression of E-cadherin was reduced by shRNA and ZEB1 was expressed in SUM149 IBC tumor cells. Numerous EMT-related genes were upregulated with loss of E-cadherin and gain of ZEB1, including N-cadherin and vimentin. However, there were marginal differences in the in vitro parameters of proliferation, Matrigel invasion and anchorage independent growth in soft agar between SUM149-shECad or SUM149-ZEB1 clones and their respective vector control cells. The loss of E-cadherin and gain of ZEB1 altered the morphology of SUM149 cells when cultured under low adherence conditions permissive for the enrichment of CSC, exhibiting a reversion in grape-like morphology to more well defined spheres, which was accompanied by increased clonogenicity in both SUM149-shECad and SUM149-ZEB1 cells. The loss of E-cadherin and the gain of ZEB1 significantly inhibited tumor growth of cells injected in the mammary fat pad of NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ mice. Tumor volume at 56 days for E-cadherin vector control cells was 771.9 mm3 +/− 185.6 compared to shECadherin tumors, which was 13.6 mm3 +/− 7.2. Tumor volume of ZEB1 vector control tumors was 346.1 mm3 +/− 96 compared to volume of ZEB1 expressing tumors, which was 21.5 mm3 +/− 7.2.Conclusions: E-cadherin with lack of ZEB1 expression in IBC is consistent with a mesenchymal-epithelial transition (MET), consistent with the retention of the epithelial phenotype while maintaining a program of rapid metastasis and colonization of lymph nodes and distant organ sites. Furthermore, we demonstrate that the E-cadherin-ZEB1 axis is critical for the in vivo growth of IBC tumor cells. Although SUM149 cells are fully capable of undergoing an EMT process, which is under negative regulation by E-cadherin, the process of EMT does not drive in vivo tumor growth in IBC.

Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P1-02-03.