Curcumin (EMD; Gibbstown, NJ) was dissolved in ethanol at a stock concentration of 8 mM. IKK inhibitor wedelolactone (EMD) was dissolved in DMSO at 15 mM stock concentration.

Triple negative MDA-MB-468 (MDA468), HCC1937, and HCC1806 breast cancer lines, HER2-over-expressing SKBR3 cells, ER-alpha-positive MCF7, and non-transformed mammary epithelial line MCF12A were purchased from ATCC (Manassas, VA). HCC1806 cells were maintained in RPMI with 5% fetal calf serum (FCS); MCF12A cells were maintained in DMEM/F12 supplemented with 5% horse serum, 20 ng/mL EGF, 10 μg/mL insulin, and 0.5 μg/mL hydrocortisone; all other lines were maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% FCS; all cell lines were maintained on 1% penicillin/streptomycin and incubated at 37 °C with 5% CO₂ in a humidified incubator.

Cells were treated with two-fold serial dilutions of curcumin for 72 hours (h), at which point cell survival was measured by trypan blue exclusion. Control cultures were treated with ethanol corresponding to the highest dose of curcumin, since curcumin is dissolved in ethanol. In addition, ethanol was added to lower dose curcumin treatment groups to make up the difference in volume of ethanol between the highest curcumin treatment group and the lower dose groups. Experiments were done at least in duplicate and performed at least twice. Cell viability is expressed as a percentage of control ethanol-treated cells per individual cell line; error bars represent standard deviation between replicates.

Cells were lysed in RIPA buffer (Cell Signaling; Danvers, MA), which includes 0.1% NP40, supplemented with protease and phosphatase inhibitor cocktails (Sigma; St. Louis, MO). Total protein extracts (50 μg) were run on SDS-PAGE and immunoblotted using the following antibodies overnight at the indicated dilutions: from EMD, HER2 (erbb2/neu) monoclonal 3B5 used at 1:1000; from Cell Signaling, p-p65 NF-kB monoclonal (93H1) used at 1:250, total p65 NF-kB monoclonal (C22B4) used at 1:250, PARP polyclonal used at 1:200; from Bethyl Labs (Montgomery, TX), p-S1189 and p-S1280 BRCA1 polyclonals used at 1:200 each; from Santa Cruz Biotechnology (Santa Cruz, CA), BRCA1 (C-20) and p-S988 BRCA1 polyclonals used at 1:200 each, ER alpha (G-20), survivin monoclonal D-8 used at 1:500; and from Sigma, β-actin monoclonal used at 1:20,000. Secondary antibodies were chosen according to the species of origin of the primary antibody. Protein bands were detected using the Odyssey Imaging System (Li-Cor Biosciences; Lincoln, NB).

Cells were plated in duplicate at 15 × 10⁵ in 6-well dishes in 1ml of matrigel (BD Biosciences; San Jose, CA) diluted 3:1 (media:matrigel). The matrigel cell suspension was allowed to solidify for at least 2 h at 37 °C. Once the matrigel solidified, 2 ml of media containing the ethanol control or curcumin (5 or 15 uM) was added to each well. The cells were maintained for approximately 2 weeks, during which media containing either control or curcumin was replenished twice a week. Photographs were taken with an Olympus IX50 inverted microscope at 4X magnification. To quantify the cell number, matrigel was digested using dispase (BD Biosciences). Briefly, the media was removed from each well and 2 mL of dispase was added per well and incubated at 37 °C for 1 h. Each 3 mL sample was then transferred to a centrifuge tube and 10 mM EDTA was added to stop the enzymatic activity of dispase. Each sample was centrifuged at 1000 rpm for 5 minutes (min) and washed 3 times with phosphate-buffered saline (PBS). Cells were then counted by trypan blue exclusion.

Monolayers of 750,000 HCC1806 cells were plated and grown in a 6-well cell culture dish. Using a p200 pipette tip, confluent cell cultures were scratched down the center. After scratching, the cells were treated with ethanol or 15 μM curcumin. Cell cultures were photographed (Olympus IX50 microscope, 4X magnification) at 0 and 24 h time points.

Cells (100,000) were plated per chamber in 4-well chamber slides in 500 μl media and incubated overnight. The next day cells were treated with ethanol or curcumin for 24 h. After washing in PBS, cells were fixed with 4% paraformaldehyde and 0.2% gluteraldehyde for 20 min at room temperature with gentle shaking. After washing, cells were permeabilized with 0.5% Triton X for 10 min at room temperature with gentle shaking. Cells were washed with PBS followed by blocking for 15 min with 5% normal goat serum at room temperature. After washing, cells were incubated overnight with the primary antibody (1:100) in 5% NGS in PBS at 4 °C. The next day cells were washed 6 times, 5 min each time with PBS and incubated for 1 hour at room temperature in a light-protected container with secondary antibody (1:250 in 5% NGS) that was chosen according to the species of origin of the primary antibody. The cells were washed 6 times for 5 min with PBS. The chamber was removed and one drop of the mounting medium which contains DAPI (Vectashield, Vector Laboratories, Inc. Burlingame, CA) was added then sealed with a coverslip. The slide was dried for 2 h in the dark at room temperature. Slides are stored at 4 °C. Photographs were taken using Zeiss Axioplan 2 Upright Microscope. Primary antibodies were BRCA1 (C-20) and phospho-gamma H2Ax polyclonal (both from Santa Cruz).

HCC1806 cells were plated in 100-mm dishes and treated with ethanol or 10 μM curcumin for 6 h or 24 h. After incubation, the cells were lysed for nuclear proteins with Nuclear Extraction Kit per manufacturer protocol (Cayman Chemical, Ann Arbor, MI). Nuclear extracts were used to detect DNA binding activity of NF-kB by using NF-kB (p65) Transcription Factor Assay Kit (Cayman Chemical). Briefly, 10 μl of the nuclear extract was added to the wells coated with a consensus dsDNA sequence in duplicates together with 90 ul complete transcription factor buffer and incubated overnight at 4 °C. Blank wells, positive control (TNF alpha-stimulated HeLa cell nuclear extract provided with kit), and non-specific binding (provided with kit) samples were also included on the plate. The next day the wells were washed 5 times and incubated with the NF-kB primary antibody (except the blank wells) for 1 h at room temperature. Wells were washed and incubated with HRP-conjugated secondary antibody (except the blank wells) for 1 h at room temperature. After washing, 100 μl of developing solution was added and incubated for 45 min at room temperature with gentle agitation followed by addition of 100 μl stop solution to each well. The absorbance was read at 450 nm. The reading for nonspecific binding was subtracted from each treatment, and the results were normalized to the nuclear extract concentration. Fold change of each sample relative to the average of untreated samples was determined.

Results were analyzed using a two-tailed Student’s t test to assess statistical significance. Values of p < 0.05 were considered statistically significant.

ER-alpha and HER2 expression status of HCC1937, HCC1806, and MDA468 cell lines was assessed by immunoblot, using MCF7 cells as a positive control for ER-alpha and SKBR3 cells as a positive control for HER2 (Fig. 1). Relative to SKBR3 cells, which harbor amplification of the her2 gene with subsequent over-expression of the HER2 protein, HCC1937, HCC1806, and MDA468 express low levels of HER2. In addition, these three lines do not show expression of ER-alpha, similar to SKBR3 which are known to be ER-alpha-negative, and compared to MCF7 cells, which are ER-alpha-positive. PR expression was previously reported as being negative in each of these lines, and other reports have confirmed the ER-negative and low HER2 expression status of these lines.26–29 Thus, HCC1937, HCC1806, and MDA468 cells represent in vitro models of triple negative breast cancer.

BRCA1 dysfunction is often observed in TNBC. BRCA1 function is regulated in part by phosphorylation and in part by cellular localization. Since other dietary polyphenols including resveratrol and indole-3-carbinol have been shown to induce expression of the BRCA1 protein in breast cancer cells,30,31 we examined the effect of curcumin on BRCA1 in TNBCs. Total BRCA1 protein expression was induced in MDA468 and HCC1806 cells within 6 h of treatment with 10 μM curcumin (Fig. 2A). Phosphorylation at serine 988, a chk2 kinase-specific phosphorylation site on BRCA1,32 was also increased in MDA468 and HCC1806 cells within 6 h of curcumin exposure. In contrast, curcumin did not alter expression or phosphorylation of BRCA1 in MCF12A non-transformed mammary epithelial cells. We examined two additional phosphorylation sites using MDA468 as a model of TNBC. Serine 1189 and serine 1280 are ATM kinase phosphorylation sites on BRCA1.32 Curcumin induced a transient increase in phosphorylation at both sites, occurring within 6 h and back to baseline again by 24 h (Fig. 2B).

Many of the molecular effects of curcumin have been attributed to its ability to potently inhibit the NF-kB transcription factor. We confirmed that curcumin inhibits NF-kB in triple negative breast cancer cells using the HCC1806 line as a model. Curcumin inhibited NF-kB transcription factor function (Fig. 3A) and phosphorylation of p65 NF-kB (Fig. 3B). Next, we compared an IKK inhibitor (wedelolactone) to curcumin for effects on BRCA1 phosphorylation and expression to determine whether curcumin-mediated changes in BRCA1 expression and phosphorylation may be due to IKK inhibition. IKK is an upstream kinase of NF-kB that inhibits NF-kB function; curcumin is thought to inhibit NF-kB via IKK activation. Wedelolactone did not induce phosphorylation or total expression of BRCA1 in contrast to curcumin (Fig. 3C). Thus, since the IKK inhibitor wedelolactone did not induce modulation of the BRCA1 protein, curcumin-mediated changes in BRCA1 may be independent of IKK inhibition.

Phosphorylation of ATM/chk2-specific sites on BRCA1 suggests that curcumin may be inducing DNA damage in TNBC cells. To test this hypothesis, we examined phosphorylation of serine 139 on histone H2Ax (gamma-H2Ax), which occurs in response to DNA double-strand breaks. Immunofluorescence showed gamma-H2Ax (GFP) foci in curcumin-treated MDA468 cells but not in control ethanol-treated cells (Fig. 4A). Collectively, these results indicate that curcumin induces DNA damage in triple negative breast cancer cells, which is associated with phosphorylation and expression of the BRCA1 DNA repair protein.

Nuclear localization of BRCA1 is necessary for activation of its transcription factor and DNA repair activity. In response to a DNA damaging agent, cytoplasmic retention of BRCA1 may occur, preventing DNA repair and promoting apoptosis. Immunofluorescence showed that BRCA1 is expressed in both the cytoplasm and nucleus in 80%–100% of control (ethanol-treated) HCC1806 and MDA468 cells (Figs. 4B and 4D). However, in the presence of curcumin (10 μM), the percentage of cells showing only cytoplasmic localization of BRCA1 is increased by 3-fold for MDA468 cells and 6-fold for HCC1806 cells. Both lines showed at least 60% of cells with cytoplasmic sequestration of BRCA1 protein upon treatment with curcumin, a statistically significant (p ≤ 0.001) change versus the control cells (Fig. 4D). On the other hand, 60% of control HCC1937 cells showed BRCA1 localized to the cytoplasm, and curcumin exposure did not change that percentage (Figs. 4C and 4D). Similarly, MCF12A cells did not show statistically significant changes in cytoplasmic localization of BRCA1 in response to curcumin. Thus, BRCA1 localization is shifted to the cytoplasm upon curcumin treatment in the triple negative cell lines MDA468 and HCC1806, which have wild-type brca1, but not in the mutant brca1 cell line HCC1937, nor in non-transformed mammary epithelial MCF12A cells. These results indicate that curcumin promotes cytoplasmic retention of BRCA1 in triple negative breast cancer cells that have functional BRCA1.

Cytoplasmic retention of DNA repair protein BRCA1 suggests that TNBC cells MDA468 and HCC1806 may be unable to repair curcumin-mediated DNA damage and may undergo apoptosis in response to curcumin exposure. In addition, lack of functional BRCA1 suggests that HCC1937 cells may have an impaired ability to repair DNA damage, and may undergo apoptosis upon treatment with a DNA damaging agent. We examined the biological response of TNBC cell lines to curcumin using trypan blue exclusion viability assays (Fig. 5A). Curcumin induced dose-dependent cell death within 72 h of treatment in all three TNBC lines. In contrast to TNBC cells, the non-transformed mammary epithelial line MCF12A did not display dose-dependent cell death in response to curcumin. Statistical analysis showed that each of the three TNBC lines showed significantly higher response to curcumin at all doses examined (p < 0.005 at 10 μM, p < 0.05 at 5 μM and 20 μM) in comparison to MCF12A cells. Non-TNBC lines MCF7 (ER-positive), SKBR3 (HER2-positive), and BT474 (ER-positive and HER2-positive) also respond to curcumin in a dose-dependent manner. However, the TNBC lines HCC1806 and HCC1937 showed statistically significant (p ≤ 0.007) higher responses to 10 μM curcumin in comparison to non-TNBC lines (Fig. 5B). MDA468 cells showed a trend toward being more sensitive than MCF7, SKBR3, and BT474, but this was not statistically significant (p = 0.16, p = 0.14, p = 0.06, respectively). These dose-response profiles indicate that curcumin induces cell death in breast cancer cells, with triple negative breast cancer cells showing a trend toward increased sensitivity versus ER-positive/HER2-over-expressing cells.

To further examine if curcumin induces apoptosis of TNBC cells, cleavage of poly (ADP-ribose) polymerase (PARP) (Fig. 5C) and expression of anti-apoptotic protein survivin (Fig. 5D) in response to curcumin were measured in MDA468 and HCC1806 cells by immunoblotting. During apoptosis, the full-length 116-kDa PARP protein is cleaved in a caspase-dependent manner into an 89-kDa fragment. Curcumin induced cleavage of PARP at 5 μM and 15 μM doses in both MDA468 and HCC1806 cells. In contrast, PARP cleavage was not observed in MCF12A cells, consistent with dose-response assays indicating that curcumin does not induce death of non-transformed MCF12A cells. In addition, curcumin (10 μM) reduced expression of survivin in MDA468 and HCC1806 cells within 24 h, consistent with induction of apoptosis.

We next evaluated the effect of curcumin on anchorage-independent growth of TNBCs (Fig. 6A). In comparison to ethanol-treated control cultures, 15 μM curcumin resulted in statistically significant prevention of anchorage-independent growth of all TNBC cell lines (Fig. 6B). To assess the ability of curcumin to prevent migration of TNBCs, we performed in vitro wound healing or “scratch” assays on HCC1806 (Fig. 6C) cells. (MDA468 and HCC1937 cell lines do not migrate well in vitro, and thus were not used for this assay). Curcumin (15 μM) inhibited HCC1806 cell migration in comparison to ethanol-treated control cells, which migrated to close wounds almost completely within 24 h. Thus, curcumin promotes apoptosis and blocks anchorage-independent growth and migration of triple negative breast cancer cells.