Inflammatory breast cancer (IBC) is the most aggressive form of locally advanced breast cancer (LABC). Patients with IBC commonly present with skin metastasis, which are observed microscopically as tumor emboli within dermal lymphatics. These metastatic tumor cells aberrantly overexpress E-cadherin and exhibit the ability to undergo self-renewal and are highly invasive. There are no therapeutics yet identified that target the structure and functions of IBC tumor emboli. The present studies evaluated the effects of the pan-histone deacetylase (HDAC) inhibitor suberoylanilide hydroxamic acid (SAHA) using IBC tumor spheroids derived from established IBC cell lines and tumor spheroids derived from pleural effusion (PE) aspirates of patients with IBC and LABC, designated as PE-IBC and PE-LABC.
Methods used are as follows: culture of IBC cells from clonal density single cells in low adherence culture conditions that promote formation of IBC tumor spheroids; clonogenic assays; cell fractionation and Western blotting; confocal microscropy; and modified Boyden chamber invasion assays.
SAHA inhibited self-renewal of IBC tumor spheroids from established IBC cell lines and PE-IBC and PE-LABC, as assessed by decreased clonogenic growth. SAHA blocked homotypic aggregation of the cells that comprised the IBC tumor spheroids leading to loss of their 3-dimensional (3D) structure, which was associated with a change in location of E-cadherin protein from the plasma membrane in untreated IBC tumor spheroids to the cytoplasm of cells within IBC tumor spheroids with SAHA treatment. In addition, SAHA blocked the robust invasion exhibited by IBC tumor spheroids of established cell lines as well as by tumor spheroids derived from PE-IBC and PE-LABC.
Inflammatory breast cancer (IBC) is the most aggressive variant of this disease, with a 5 year overall survival rate for IBC patients that is significantly lower compared with the overall survival rate for patients with non-IBC locally advanced breast cancer (LABC).1 Although multimodality therapy, including neoadjuvant chemotherapy, followed by surgical resection and radiation, has been used to treat IBC patients older than 30 years, there has been no change in overall survival of IBC patients during this time period.2 There is an urgent need to identify therapeutic agents that target specific characteristics that are associated with the highly aggressive phenotype of IBC.
One of the signatures of IBC is the presence of invasion of tumor cells into the lymphatics of the dermis.1 IBC tumor emboli within the dermal lymphatics aberrantly overexpress the calcium-dependent adhesion molecule E-cadherin. This protein, which is located at the adherens junctions on the plasma membrane, mediates homotypic aggregation, which confers the 3-dimensional (3D) structure of the tumor emboli.3 E-cadherin has been associated with a normal phenotype, with loss of E-cadherin reported to commonly occur during the epithelial-mesenchymal transition (EMT). However, a recent study suggests that tumors with a highly malignant phenotype expression of E-cadherin, which occurs in conjunction with its binding partner p120 catenin and small GTPases, results in increased motility and invasion.4 These observations are consistent with E-cadherin overexpression on the surface of highly invasive IBC tumor emboli. In addition to E-cadherin expression, IBC tumor emboli have been shown to possess characteristics consistent with this tumor type being enriched for cells with a tumor initiating ie, cancer stem cell phenotype, including the ability to undergo self-renewal.5
To date, no therapeutic agents have been identified that target the characteristics of IBC, such as homotypic aggregation of IBC tumor emboli and self-renewal. Recent evidence suggests that histone deacetylase enzymes (HDAC) enzymes, in tandem with histone acetylases (HAT) enzymes, control reversible chromatin remodeling, which is a key epigenetic mechanism that regulates transcription of genes associated with multiple activities of normal and malignant cells, including self-renewal, survival and resistance to apoptosis, cell cycle progression, aggregation, motility, invasion, angiogenesis, and metastasis.6, 7
Several HDAC inhibitors have been developed that inhibit the activity of multiple HDAC enzymes. Suberoylanilide hydroxamic acid (SAHA; Vorinostat; Zolinza, Merck & Co., Inc.) is the first pan-HDAC inhibitor to be approved by the FDA.8, 9 Although the first indication for SAHA was for the treatment of patients with CTCL, recent studies suggest that this pan-HDAC inhibitor may have activity in solid tumors, including metastatic breast cancer.10 Preclinical studies and clinical trials are underway to evaluate both pan-HDAC inhibitors and selective HDAC inhibitors for their utility in treatment of multiple types of solid tumors as well as hematological disorders.
Not only has there been a lack of progress in development of effective treatments for IBC, but there are also few cell lines available with which to study IBC. Only 2 cell lines are widely available for study. The SUM149 IBC cell line was developed from pleural effusion cells of an IBC patient and is the most studied of all of the IBC cell lines.11, 12 SUM190 IBC cells were developed from the primary tumor of an IBC patient. Both of these IBC cell lines are used in the present study. Because of the lack of cell lines with which to study IBC and LABC, the present study also used tumor cells isolated from pleural effusion of a patient with IBC and a patient with LABC, as a comparison to the response of the established IBC cell lines to SAHA. By using a combination of established IBC cell lines and tumor cells isolated from patients with metastatic IBC and LABC, the present studies evaluated the effects of SAHA on specific characteristics associated with the aggressive phenotype of IBC.
MATERIALS AND METHODS
Cell Lines and Conditions
The human MCF-7 breast cancer cell line was obtained from American Type Cell Culture (Manassas, Va). MCF-7 is an estrogen receptor positive (ER+) breast cancer cell line used as a control cell line in these studies. MCF-7 cells were cultured in Dulbecco modified Eagles medium (DMEM/F12; Invitrogen, Carlsbad, Calif) and supplemented with 10% heat-inactivated fetal bovine serum (FBS; Invitrogen). The SUM149 and SUM190 cell lines were provided by Dr. Stephen Ethier (Asterand, Detroit, Mich).11, 12 SUM149 and SUM190 cells were maintained as 2-dimensional (2D) adherent cultures in Ham F12 Nutrient Mixture (Invitrogen) and supplemented with 10% FBS, insulin (1 mg/mL; Sigma-Aldrich, St. Louis, Mo) and hydrocortisone (1 mg/mL; Sigma-Aldrich). For low adherence culture conditions, which promote formation of tumor spheroids and self-renewal, SUM149 and SUM190 cells as well as pleural effusion (PE)-IBC and PE-LABC cells were maintained in serum-free mammary epithelial growth medium (MEGM, BioWhittaker) supplemented with B27 (Invitrogen), 20 ng/mL epidermal growth factor (EGF), 40 ng/mL bFGF (BD Biosciences), and 4 ug/mL heparin (Sigma).13, 14 All cell lines were maintained at 37°C under 5% CO2 in humidified incubators.
Assessment of Self-Renewal as Measured by Clonogenic Growth
Single cell suspensions were plated into ultra low attachments plates or flasks (Corning) and maintained as described above. Tumor spheroids were treated with either dimethyl sulfoxide (DMSO) as the solvent control or with SAHA at concentrations of 0.25 μM, 0.5, 2.5, 5.0, 10.0, 25.0, or 50.0 μM, with each concentration run in triplicate. After 10 days in culture, 0.5% thiazolyl blue tetrazolium bromide (MTT) was added to each well to allow visualization of the tumor spheroids and samples were returned to the incubator for 1 hour. Formation of tumor spheroids was measured with an Optronix Scanner (Oxford, UK) and data was analyzed using GelCount software. The IC50 for SAHA was calculated using nonlinear regression analysis of average tumor spheroid formation. The concentration of SAHA that inhibited 50% of spheroid formation, designated as the IC50 values, are represented from data obtained in 3 separate experiments for all cell lines and pleural effusion cells evaluated, with the Pearson correlation coefficients shown in Table 1.
Table 1. SAHA Inhibits Clonogenic Growth of IBC Tumor Spheroids
SAHA significantly inhibits clonogenic growth of tumor spheroids from established IBC cell lines SUM190 and SUM149 and from tumor spheroids obtained from PE-IBC and PE-LABC. IC50 concentrations for the inhibitory effect of SAHA ranged from 0.17 μM for SUM149 to 3.31 μM for SUM190.
SAHA indicates suberoylanilide hydroxamic acid; IBC, inflammatory breast cancer; PE, pleural effusion; LABC, locally advanced breast cancer.
E-Cadherin Antibody Staining of IBC Spheroids and Photographic Assessment of Homotypic Aggregation of IBC Tumor Cells
Confocal microscopy was used to define the location of E-cadherin on IBC tumor spheroids. Spheroids were treated with SAHA (5 μM, 6 hours, 24 hours), and slides were prepared by centrifuging spheroids onto slides at 500 rpm for 3 minutes using a Shandon cytospin. Slides were rinsed with phosphate buffered saline (PBS) at room temperature, fixed in 4% paraformaldehyde in PBS for 15 minutes at room temperature, followed by additional washes in PBS. Cells were permeabilized with 0.1% Triton-X 100 in PBS for 5 minutes, followed by incubation in 5% goat serum in PBS for 1 hour to block nonspecific antibody binding. Slides were then incubated with E-cadherin polyclonal antibody (Cell Signaling, Danvers, Mass) in PBS for 18 hours at 4°C. Nonspecific staining was determined by incubating slides with isotypic control antibody as the staining control. Slides were then washed in PBS 3 times for 10 minutes each, and incubated with goat-antirabbit secondary antibody conjugated with Alex-Fluor 568 (Invitrogen, Carlsbad, Calif) in PBS at a 1:400 dilution for 2 hours at room temperature. For nuclear staining, slides were then incubated with TO-PRO-3 (Invitrogen) in PBS at a dilution of 1:300 for 15 minutes. Slides were then washed in PBS 3 times for 10 minutes each, followed by placing cover slips on slides using antifade mounting medium (Vector Laboratories, Burlingame, Calif). Images were captured using a Zeiss LSM 510 confocal microscope. Representative confocal images are shown to allow visualization of the location of E-cadherin protein on SUM149 spheroids and to observe the loss of homotypic aggregation induced by exposure of SUM149 spheroids to SAHA (Figs. 3 A-F).
Western Blot Analysis of Effects of SAHA on Amount and Location of E-Cadherin Protein in IBC Tumor Spheroids
SUM149 and SUM190 cells were cultured either under adherent conditions in Ham F12 medium with 10% FBS or under low adherence conditions in MEGM with growth factors and supplements. For analysis of total E-cadherin protein, cells were treated with 5 μM SAHA for 24 hours. After treatment, cells were isolated and spun down to isolate cell pellets, which were then lysed in 1% M-PER lysis buffer (Thermo Fisher Scientific, Rockford, Ill). For analysis of E-cadherin protein in different cell fractions, cells were cultured under low adherence conditions, treated with SAHA at 5 μM for 24 hours. Cells were isolated and centrifuged to obtain cell pellets, with removal of the supernatants to obtain pellets as dry as possible. Pellets were placed in buffers to isolate cytoplasmic or nuclear fractions using the nuclear and cytoplasmic extraction kit (Pierce Biochemicals, Rockford, Ill) according to manufacturer's instructions. Equal amounts of protein (50-100 μg), as determined by total protein assay (Bio-Rad Laboratories, Hercules, Calif), were then separated using 10% polyacrylamide gels (Bio-Rad). Proteins were transferred to nitrocellulose membranes (GE Healthcare, Piscataway, NJ), blocked for nonspecific binding using a buffer containing 1X PBS, 0.1% Tween-20, and 5% milk and then probed with E-cadherin antibodies (Cell Signaling, Danvers, Mass), followed by incubation with antirabbit IgG horseradish peroxidase-conjugated secondary antibodies (GE Healthcare). For analysis of total protein in whole cell lysates and cytoplasm fractions, E-cadherin protein was normalized to β-actin, used as a loading control. For analysis of E-cadherin protein in nuclear fractions, Histone 3 was used as a loading control. Protein bands were observed by Chemiglow enhanced chemiluminescence system (Alpha Imager, San Leandro, Calif). Images were scanned and quantified using an Alpha Innotech densitometer with the Alpha Imager application program. Experiments were repeated 3 times and representative Western blots and densitometric analysis are shown.
Assessment of Invasion Through an Artificial Matrigel Basement Membrane
The effects of SAHA on in vitro invasion of SUM149 and SUM190 IBC spheroids and tumor spheroids from PE-IBC and PE-LABC through a Matrigel basement membrane was determined based on the number of cells that invaded through transwell inserts coated with the artificial basement membrane, Matrigel. Tumor spheroids were trypsinized, resuspended in media, and counted. Plates (6-well) compatible with transwell inserts with 8 μm pore-sized polycarbonate filters (Fisher Scientific) were coated with Matrigel in cold serum-free DMEM/F12 at a final concentration of 0.7 mg/mL and placed at room temperature for 40 minutes. Cells (in 500 μL serum-free medium) were added into the transwell inserts and incubated for 72 hours in absence or presence of SAHA. As a control, 10% FBS was used to evaluate the baseline extent of invasion of tumor spheroids. The lower chambers were filled with 2 mL of media. After incubation, noninvasive cells on the upper surface of the filter were removed with cotton swabs. Cells that had passed through the pores onto the lower side of the filter were fixed, stained with Hema-3 stain (Fisher Scientific), and quantified. The experiments were performed in triplicate and repeated twice. Two-tailed Student t tests were performed to evaluate the effects of SAHA on invasion of IBC tumor spheroids.
SAHA Inhibited Clonogenicity of IBC Tumor Spheroids
The effects of the pan-HDAC inhibitor SAHA on self-renewal of IBC cells cultured under low adherence conditions, which promote formation of tumor spheres, was assessed. For these studies, clonogenicity of tumor spheroids from established IBC cell lines as well as PE-IBC, and PE-LABC was evaluated. SAHA induced a dose dependent inhibition of self-renewal of SUM149 and SUM190 IBC tumor spheroids as well as inhibited self-renewal of freshly isolated PE-IBC and PE-LABC. As shown in Table 1, the concentrations of SAHA that inhibited 50% of clonogenic growth (IC50) of tumor spheroids ranged from 0.17 μM for SUM149 to 3.31 μM for SUM190, with the IC50 of PE-IBC and PE-LABC at doses between those observed for the SUM190 tumor spheroids. The IC50 doses of SAHA for all of the IBC cells was within the clinically achievable range of 2.5 μM SAHA.
SAHA Inhibited Homotypic Aggregation and 3D Integrity of IBC Tumor Spheroids
Both SUM149 and SUM190 cells form tumor spheroids when placed in low adherence conditions in defined medium. Although SUM190 cells form very tight tumor spheres, SUM149 cells form tumor spheroids that are not as tightly packed as the SUM190 cells (Fig. 1 A and B). Phase contrast microscopy was used to observe the dose dependent effects of SAHA (2.5 μM-10 μM; 24 hours) on homotypic aggregation of SUM149 and SUM190 IBC tumor spheroids, revealing that SAHA induced a loss of integrity of both SUM149 and SUM190 tumor spheroids (Figs. 1 A and B). Similar dose and time dependent results were observed with treatment of PE-IBC with SAHA (data not shown).
SAHA Altered the Location of E-Cadherin Protein in IBC Tumor Spheroids
SUM149 and SUM190 cells produce abundant E-cadherin protein when cultured under both 2D adherent conditions and in 3D low adherence culture conditions, which promote formation of tumor spheres (Fig. 2 A and B). When cultured as adherent cells, SUM149 cells produce approximately a 2-fold greater amount of E-cadherin protein, which included a higher molecular weight glycosylation product, which was not detected in SUM149 tumor spheroids cultured under low adherence conditions (Fig. 2 A and B). SUM190 cells produced equivalent amounts of E-cadherin protein regardless of the culture conditions. When cultured as tumor spheroids, both SUM149 and SUM190 cells produced E-cadherin protein, with no change in total E-cadherin protein when treated with SAHA (Fig. 2 C). Cell fractionation studies combined with Western blotting and densitometry revealed that although SAHA did not alter the total E-cadherin protein produced by IBC tumor spheroids, SAHA treatment resulted in a change in the location of E-cadherin protein on IBC tumor spheroids. In untreated IBC tumor spheroids, E-cadherin was located on the plasma membrane, which was translocated into the cytoplasm by SAHA treatment (Fig. 2 E and F).
SAHA Altered Location of E-Cadherin Protein and Inhibits Homotypic Aggregation of IBC Spheroids
Confocal microscopy was used to observe the distinct pattern of E-cadherin protein localized to the adherens junctions of cells within SUM149 tumor spheroids. In untreated SUM190 spheroids, E-cadherin protein was located at the surface of the plasma membrane at the point of cell:cell contact between cells within tumor spheroids (Fig. 3 A-C). Treatment of SUM149 tumor spheroids with 5 μM SAHA for 6 hours induced a loss of E-cadherin protein at the adherens junctions, with E-cadherin protein visible within the cytoplasm of IBC tumor cells that had been detached from the spheroid structure (Fig. 3 D-F). The SAHA-induced alteration in E-cadherin localization at the plasma membrane occurred simultaneously with the diminished homotypic aggregation of cells within the spheres, resulting in loss of integrity of the 3D tumor spheroid structures (Figs. 3 D-F). The results of confocal imaging studies are consistent with results shown in Figures 1 and Figures 2 which demonstrated that SAHA did not alter the total amount of E-cadherin protein in IBC tumor spheroids but mediated a translocation of protein to the cytoplasm of IBC tumor spheroids, which was associated with the loss of homotypic aggregation and integrity of IBC tumor spheres.
IBC Tumor Spheroids Exhibit Robust Invasion Across an Artificial Basement Membrane that is Blocked by SAHA
SUM149 cells cultured as either adherent cells or as tumor spheroids exhibit the ability to invade across an artificial basement membrane. In contrast, SUM190 cells are not invasive when cultured under 2D adherent conditions but acquire this characteristic when cultured under low adherence conditions. SAHA significantly (P < .05) inhibited invasion of SUM149 and SUM190 tumor spheroids as well as invasion exhibited by PE-IBC and PE-LABC (Fig. 4).
IBC has a propensity to invade into the lymphatic system of the dermis1 and to form aggregates of cells, defined as tumor emboli, which exhibit homotypic aggregation that is mediated through the aberrant overexpression of E-cadherin.3, 5 In addition, IBC tumor cells have been shown to express surface characteristics consistent with the enrichment for cells with a “tumor initiating” phenotype, ie, cancer stem cells that are capable of self-renewal and multipotency.5 There are no therapeutic agents yet identified that effectively inhibit self-renewal and the tight homotypic aggregation that are characteristics of IBC tumor emboli.
The present studies used established IBC cell lines SUM190, SUM149, as well as used tumor cells isolated from pleural effusion aspirates from a patient with IBC and a patient with LABC, designated as PE-IBC and PE-LABC, respectively, to evaluate the effects of the pan-HDAC inhibitor SAHA, on self-renewal of IBC tumor spheroids and on their tight homotypic aggregation. SAHA significantly inhibited self-renewal of IBC tumor spheroids formed by established IBC cell lines SUM149 and SUM190 and by PE-IBC and PE-LABC. The inhibitory effects of SAHA on homotypic aggregation were apparent visually at a microscopic level using phase contrast microscopy and appeared as a loss of integrity of the 3D structure of IBC spheroids. On the basis of the known role of E-cadherin in mediating homotypic aggregation, further studies were performed to determine the effects of SAHA on E-cadherin protein production by IBC tumor spheroids.
The present results confirm previous reports that IBC cells express high levels of E-cadherin.3, 5 Previous studies using anti-E-cadherin antibodies and small interfering (si) ribonucleic acid (RNA) approaches to knockdown E-cadherin have demonstrated that the result of effectively blocking the homeotypic aggregation of IBC tumor emboli is the inhibition of IBC metastasis.5 Interestingly, although SAHA did not alter the total amount of E-cadherin protein, it altered its location, leading to a loss of homotypic aggregation of IBC tumor spheroids. SAHA is among the first therapeutic agents that have been shown to effectively alter E-cadherin function in IBC. Although E-cadherin has been reported to commonly be lost during the process of EMT, which is associated with tumor progression and metastasis, IBC is a notable exception in that it is overexpressed during IBC invasion into the dermal lymphatics.3, 5 The specific role of E-cadherin in IBC metastasis and the mechanisms of its regulation are currently unclear and represents an important avenue for further study. Future studies will elucidate the associations between aberrant E-cadherin expression in IBC, and its association between its cytoplasmic binding partner p120 catenin, small GTPases, and beta catenin signaling pathway, which may together be critical to mediating the rapid invasion observed in IBC.
Taken together, the present studies demonstrate that cells derived from established SUM190 and SUM149 IBC cell lines as well as cells isolated from pleural effusion of patients with metastatic IBC and LABC exhibit self-renewal as assessed by clonogenicity, which is consistent with an enriched subpopulation with characteristics of cancer stem cells. IBC tumor spheroids serve as a good in vitro surrogate model with which the peculiar characteristics exhibited by IBC tumors can be studied, including the aberrant overexpression of E-cadherin and the homotypic aggregation of IBC cells that exhibit robust invasive activity. By using approaches such as those taken in the present experiments, it is possible to identify potential therapeutic targets that can be matched by the availability of FDA-approved therapeutic agents, which can be immediately evaluated for their effects on characteristics that define the aggressive phenotype of IBC. In addition, the present results are the first to demonstrate the potential utility of agents that are capable of epigenetic modulation of IBC tumors as a strategy to target the molecular signatures that define IBC as the most aggressive form of locally advanced breast cancer.
The technical assistance of Kimberly Boley is gratefully acknowledged.
CONFLICT OF INTEREST DISCLOSURES
These studies were supported in part by American Airlines-Komen Foundation Promise Grant KGO81287 (FMR, MC), NIH NCI R01 CA-138239-02 (WW, MC), and the State of Texas Grant for The Rare and Aggressive Breast Cancer Research Program (MC).