Alcohol promotes breast cancer cell invasion by regulating the Nm23-ITGA5 pathway

T47D, MCF-7 and MDA-MB-231 breast cancer cells were purchased from American Type Culture Collection (Manassas, VA, USA). Cells were cultured at 37°C, 5% CO2, on 75-cm³ tissue culture flasks (Becton Dickinson Labware, Franklin Lakes, NJ, USA) in Dulbecco's Modified Eagle's Medium (DMEM) supplemented with 10% inactivated fetal bovine serum (FBS) and 1% penicillin-streptomycin (Gibco, St Louis, MO, USA). The Nm23 siRNA, ITGA5 siRNA, and negative controls were purchased from Invitrogen (Carlsbad, CA, USA). pcDNA3-Nm23-H1 cDNA and the control vector were kindly provided by Dr. Patricia Steeg (National Cancer Institute, Bethesda, MD, USA). T47D cells were transfected with the above vectors and siRNAs using Lipofectamine 2000 (Invitrogen) following the manufacturer's instructions. Neomycin-resistant clones were isolated by growth in media containing 800 ug/ml G418 (Gibco, St Louis, MO, USA). Alcohol was added to the medium at concentrations of 0.1%, 0.2%, and 0.5% v/v ethanol. RNA and proteins were collected from the cells 48 hours post alcohol treatment.

The in vitro invasion studies were performed using the BD Bio-Coat Matrigel invasion assay system (Becton Dickinson Labware, Franklin Lakes, NJ, USA). To determine the ability of alcohol to affect the invasive ability of breast cancer cells, 2 × 10⁵ T47D cells were suspended in serum-free DMEM medium containing 0.1% bovine serum albumin (BSA) and placed in the upper chamber. The bottom chamber was filled with DMEM containing 10% FBS. The FBS attracted the cancer cells and triggered their migration to the underside of the membrane. Breast cancer cells that have the ability to invade secrete factors which allow them to degrade the Matrigel (e.g., matrix metalloproteinases) and migrate through the 8 μm pores to the lower chamber of the membrane. After 24 hour incubation, the membrane of the upper chamber was cleaned with cotton swabs to remove the Matrigel and the cells that did not migrate. The membrane was fixed and stained using Diff-Quik solutions (Dade-Behring, Newark, DE). Staining of cells allows their visualization and quantification using a light microscope. Five fields of adherent cells were randomly counted in each well with a Nikon Diaphot-TMD (Atlantic Lab Equipment, Salem, MA, USA) inverted microscope at 20× magnification.

Total RNA was extracted using the RNeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. Reverse transcription was performed with the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA, USA), using 2 mg of RNA for each reaction. Primer pairs were designed using Primer3 software [22] and are shown in Table 1. Real-time PCR was performed with the SYBR GreenER qPCR kit (Invitrogen, Carlsbad, CA, USA) in the Mastercycler ep Realplex Real-time PCR thermocycler (Eppendorf, Wesseling-Berzdorf, Germany). The relative expression levels of target genes were normalized to the housekeeping gene 18S rRNA. Amplification specificity was confirmed by melting curve analysis.

Cells were lysed using RIPA buffer containing 50 mM Tris (pH 7.6), 150 mM NaCl, 2 mM EDTA, 20 mM MgCl2, 1% Nonidet P40 containing protease inhibitors (1 μg/ml PMSF, 1 μg/ml aprotinin and 1 μg/ml pepstatin). Samples were incubated for 1 hour on ice with agitation and centrifuged at 12,000 × g for 20 min. Protein samples were subjected to electrophoresis on 4-12% SDS-polyacrylamide gradient gels and transferred to a PVDF membrane. Membranes were probed with anti-Nm23-H1 (BD Biosciences, San Jose, CA, USA) and anti-actin (Oncogene, Cambridge, MA, USA) antibodies. Protein-antibody complexes were detected with horseradish peroxidase-conjugated secondary antibodies (Cell Signaling Technology, Danvers, MA, USA) followed by enhanced chemiluminescence reaction. Immunoblots were quantified using ImageJ software (NIH website: http://​rsbweb.​nih.​gov/​ij/​index.​html).

Total RNA was extracted using the RNeasy Mini Kit (Qiagen, Hilden, Germany). The cDNA was prepared by reverse transcription using the RT² PCR Array First Strand kit (SA Biosciences, Frederick, MD) as recommended by the manufacturer's instructions. PCR array analysis of 84 genes related to cell-cell and cell-matrix interactions as well as human extracellular matrix and adhesion molecules (RT² Profiler™ PCR array, PAHS-013A-1, SA Biosciences, Frederick, MD, USA) was performed using the Mastercycler ep Realplex real-time PCR thermocycler (Eppendorf, Wesseling-Berzdorf, Germany). Briefly, 25 μl of PCR mixture, which contained cDNA equivalent to 1 μg RNA in SuperArray RT² qPCR Master Mix solution, was loaded in each well of the PCR array plate. PCR amplification of cDNA was performed under the following conditions: 10 min at 95°C for one cycle, 15 sec at 95°C, followed by 1 min at 60°C for 40 cycles. All mRNA Ct values for each sample [Ct (sample)] were normalized to glyceraldehyde-3-phosphate dehydrogenase [Ct (GAPDH)] in the same sample. The relative mRNA level was expressed as the value of 2^-ΔΔCt (sample).

One-way analysis of variance (ANOVA) was used to test the statistical significance of the qRT-PCR and invasion assay results (SPSS 12.0 student edition, SPSS Inc. Chicago, IL, USA). To detect statistical significance, p value was set at 0.05, and data are presented as the mean ± standard error of the mean (SEM).

To investigate the role of alcohol in cell invasive ability, human breast cancer T47D cells were treated with 0.1%, 0.2%, and 0.5% v/v ethanol for 24 hours. Previous studies have shown that alcohol exposure at these concentrations and length of time in vitro yielded biological effects seen in breast cancer patients [23, 24]. We show that alcohol treatment in vitro increased the ability of T47D cells to invade in a dose-dependent manner (Figure 1A). Treatment with 0.1%, 0.2%, and 0.5% v/v alcohol increased cell invasion by approximately two-, four-, and six-fold, respectively (Figure 1A, p < 0.05). Similar results were seen with MCF-7 and MDA-MB-231, human breast cancer cell lines with low and high, respectively, invasive potential (Figure 1B).

To investigate the possibility that alcohol may increase cellular invasive ability by inhibiting the expression of specific metastasis suppressing genes, we determined the effects of alcohol on known metastasis suppressor genes. We examined the effects of 0.5% v/v ethanol on the expression levels of Nm23, KISS1, Mkk4, RRM1, KAI1, and BRMS1 metastasis suppressor genes in vitro by qRT-PCR (Figure 2). Our results show that alcohol significantly suppressed the expression of Nm23 by approximately 50% (Figure 2, p < 0.05), suggesting that the Nm23 metastasis suppressor gene may be involved in alcohol-induced cell invasion.

To determine whether the effects of alcohol on the invasive ability of T47D cells can be blocked via Nm23, we transfected T47D cells with the pcDNA3-Nm23-H1 vector (kindly provided by Dr. Patricia Steeg at the National Cancer Institute, Bethesda, MD, USA) to overexpress Nm23. As expected, Nm23 overexpression resulted in a significant decrease in T47D cell invasion (Figure 3A, p < 0.05) while treatment of T47D control cells (transfected with an empty vector) with 0.5% v/v alcohol significantly increased cell invasive ability (Figure 3A, p < 0.05). (Note: Results from Figure 1A and 3A indicate that 0.5% v/v ethanol increased cell invasion by 600% and 50%, respectively. This difference may be attributed to the addition of G418 (Gibco, St Louis, MO, USA) in the media used for the invasion assay shown in Figure 3A. As an inhibitor of protein synthesis, addition of G418 may have led to a decline in cell proliferation over the 24 hour invasion period.) However, 0.5% v/v alcohol was unable to increase the invasive ability of T47D cells overexpressing Nm23 (Figure 3A, p > 0.05), suggesting that Nm23 expression is critical in alcohol-induced T47D breast cancer cell invasion. Nm23 protein levels are shown in Figure 3B.

To examine the downstream targets of Nm23 involved in alcohol induced cell invasion, we determined the effects of Nm23 overexpression and 0.5% v/v ethanol treatment on 84 genes associated with extracellular matrix regulation and adhesion molecules in the following groups of breast cancer cells: 1) T47D controls cells (empty vector), 2) T47D cells treated with 0.5% v/v alcohol (empty vector), 3) T47D cells overexpressing Nm23, and 4) T47D cells overexpressing Nm23 and treated with 0.5% alcohol. Results are presented in Table 2, with only the most significantly affected genes shown. Interestingly, one gene observed to be affected by alcohol and Nm23 in the opposite manner was fibronectin receptor subunit integrin alpha 5 (ITGA5). In cells overexpressing Nm23, alcohol treatment was no longer able to increase ITGA5 expression (Table 2). Additionally, alcohol exposure increased the expression of ITGA5 nine-fold; however, this effect was eliminated by the overexpression of Nm23 (Figure 4A and Table 2), suggesting that Nm23 blocked the effects of alcohol. Thus, our data suggests that the effects of alcohol on ITGA5 are Nm23-dependent.

To determine the relationship between Nm23 and ITGA5 in alcohol-treated T47D breast cancer cells, we knocked down each gene separately and in combination, using small interfering RNA (siRNA), and subsequently measured cell invasion. If alcohol increases the invasive ability of T47D cells through the down-regulation of Nm23, as suggested earlier, then down-regulation of Nm23 should increase the invasiveness of T47D cells. Indeed, results show that knock-down of Nm23 by siRNA increased the invasiveness of T47D cells and alcohol was unable to further increase the invasive ability of T47D cells significantly when Nm23 was suppressed (Figure 5A). This work is in agreement with our results in Figure 2 and provides further evidence that alcohol increases the invasiveness of T47D cells through Nm23.

To establish the relationship between alcohol, Nm23, ITGA5 and cell invasion, we knocked down ITGA5 with siRNA in T47D cancer cells and measured the ability of alcohol to affect the invasive ability of these cells. Results show that down-regulating ITGA5 significantly inhibited the ability of T47D breast cancer cells to invade (Figure 5A, p < 0.05). In agreement that decreased ITGA5 expression reduces cell invasive ability, we show that both the Nm23 overexpressing cells and the alcohol-treated Nm23 overexpressing cells have significantly reduced ITGA5 expression (Figure 4A) as well as have an overall lower cell invasive ability (Figure 3A) compared to controls. We also show that alcohol-treated Nm23 overexpressing cells have slightly higher ITGA5 levels compared to non-alcohol-treated Nm23 overexpressing cells (Figure 4A) and this translated to a slightly higher, although not statistically significant, number of invaded cells (Figure 3A). Nm23 and ITGA5 protein expression in T47D cells is shown in Figure 4B. To examine whether the Nm23-ITGA5 effects on invasion were specific to T47D cells, we exposed MCF-7 and MDA-MB-231 cells to various doses of ethanol. We show that alcohol is able to increase Nm23 and decrease ITGA5 in a dose-dependent manner (Figure 4C) and this correlated with increasing cell invasive ability (Figure 1B). Moreover, when ITGA5 was knocked down with siRNA, alcohol was unable to increase the invasion of T47D cancer cells, suggesting that ITGA5 is necessary for alcohol to increase the invasive ability of T47D cancer cells. Furthermore, in ITGA5 knocked-down cells, suppression of Nm23 by siRNA did not rescue their invasive ability (Figure 5A). Results also show that Nm23 knock-down increased ITGA5 expression; however, knockdown of ITGA5 did not affect Nm23 expression (Figure 5B), suggesting that Nm23 is an upstream factor of ITGA5. Depletion of Nm23 and ITGA5 in T47D cells following siRNA transfection is shown in Figure 5C. In summary, the above findings suggest that alcohol increases the invasive ability of breast cancer cells by down-regulating Nm23, which increases ITGA5 expression, and this elevation in ITGA5 increases the ability of breast cancer cells to invade.